i, ~ 2~9~P~'i7~
NEW DIAGNOSTIC AND TREAT~riEi~iT Mi~iODS INVOLViNG THE
cYsric FIBROSIS TRANS~iEMBRANE REGUWOR
Related Applications This application ~ a continuation-~n-part application of U$N 07/488~07, filed March 5,1990, and of USSN 07/589,295, filed September 27,1990, both copending.
Fleld of the Inventlon This invention relates to the use of recombinant DNA techniques to produce the cystic flbrosis transmembrane conductance regulator (CFTR), and in particular itrelates to new methods for detectlng CFTR and CFTR related defects and to new treatment methods therefor.
Back~round of the Invention Cystlc flbrosis (CF) is the most common fatal genetlc disease in humans (Boat etal.,1989). Based on both genetic and molecular anaiysis, a ~ene associated with CF
was recently isolated as part of 21 individual cDi~iA clones and its protein product predicted (Kerem et Q.,1989; Riordan et al.,1989: Rommens et QL,1989). USSN
488 307 descrioes the constructlon of the gene into a continuous strand and conflrmed the ~ene is responslble for CF by Introductlon of a cDNA copy of the codin~ sequence into epithellal celis from CF patients (See also Gregory et aL,1990:
Rlch et al.,1990). Wild type but not a mutant verslon of the cDNA complemented the defect In the cAMP regulated chloride channel shown previousiy to be characteristic of CF. Slrnllar concluslons were reported by others (Drumm et al.,1990).
The proteln product of the CF assoclated ~ene Is called the cystlc fibrosis transmembrane conciuctance re~3ulator (CFTR) (Rlordan et aL,1989). CFTR is a protein of approxlmateiy 1480 amlno aclds made up of two repeated elements, each 3s comprisln~ six transmembrane segments and a nucleotlde blnding domain. The two repeats are separated by a large, polar, so-called R-domain containing muitlple 1G4-9.2 ' -.
(f ~ O ~ n r A
potential phosphorylation sites. i3ased on ris predicted domain structure, CFTR is a member of a class of related proteins whlch mcludes the muHI-dru~ reststance (MDR) or P-~lycoprotein, bovine adenyl cyclase, the yeast STE6 protein as well as several bacterial amino acid transport proteins (Rlordan et aL,1989: Hyde et al.,1990).
Protelns In this ~roup, characteristlcally, are invoived In pumpin~ molecules into or out of celis.
Ci-TR is a lar~e, muitl domain, Inte~ral membrane proteln wh~h is postulated to re~ulate the outward fiow of anions from epithelial cells in response to phosphorylation 10 by cycllc AMP-dependant protein klnase or proteln klnase C (Rlordan et al.,1989:
Weish,1986: Frizel et al.,1s86 Weish and Uedtke,1986 Schoumacher et al.,1987; U et aL, 1988: Hwang et aL,1989; U et al.,1989), To Investi~ate the function of the cETr~ the mechanism by whlch mutations in 15 the CFTR ~ene cause cystic fibrosis, to deveiop potentlal therapies for cystlc flbrosls, and for many other applications, a cDNA ck~ne encodin~ the entire Ci-TR protein is necessary.
It is an aspect of the present Invention to en~ineer Ci~n? cDNA sequences 20 containin~ all of the codin~ tnformation for Ci~rR protein on a sin~le recombinant DNA
molecule which can be stably propa~ated in E. col~ and transferred to yeast, insect, plant or mammallan celis, or trans~enlc animals, for expression of wild-type Ci ll?
protein, as well as mutant to provide derivatives whlch correlate wNh the cystic flbrosis disease.
It Is another aspect to provide the crNical cDNA ~ene containln~ the correct ~ene sequence In order to provide for productlon of the Ci TR proteln.
It Is yet another aspect to enable various dla~nostic, therapeutic and protein productlon technlques related to the evaluation and treatment of cystlc flbrosiscaused by tauNy CFTiR functlon, fauNy Ci~ll? processln~ or related to the Intracellular location of Ci rR.
In addNion, a mutatlon withln the ~ene sequence encodln~ Ci TR protein has been identified In Di~iA samples from patlents wHh wNh cystic flbrosis, the most 1G4-9.2 J ?J J r~
common ~enetic disease of caucasians (Kerem et al., 1989). The mutation, which resuHs In the deletlon of the amino acid phenylalanine at posHlon 508 of the Ci TR
amino acid sequence, is assoclated wnh approximately 70~ of the cases of cystic fibrosis.
This mutation in the Ci TR ~ene resuits In the failure of an epHhelial cell chloride channel to respond to cAMP (Frizell et al.,1986; Weish, 1986; Li et al.,1988; Quinton, 1989). In airway cells, this leads to an imbalance In ion and fluid transport. It Is widely beiieved that th'ls causes abnormal mucus secretion, and uitimateiy resuits In pulmonary infection and epHhelial cell dama~e. That the chloride channel can be re~ulated by cAMP in isolated membrane patches (Li et al.,1988) suçiç~ests that at least some Ci TR is present in the apical piasma membrane and that Ci TR responds to proteln kinase A. Whether Ci TR Hseif is a re~ulanor of the membrane chloride channel or constHutes the channel ~rtseif remains controversial.
USSN 488,307, fully incorporated herein, showed that c~ is a membrane-associated ~Iycoprotein that can be phosphorylated in vitro (Gre~ory et ai.,1990).
The protein has a primary translanlon product which ml~rates wHh apparent molecular wel~ht on SDS-poiyacrylamide ~els of 130k (referred to as band A). In vaccinia virus-2 o infected, cDNA transfected HeLa celis or in reticulocyte Iysates containin~ canine pancreatlc membranes, band A is mod'lfied by ~iycosylation to yield a verslon ofapparent molecularwei~ht 135kd called band B. The use of polyclonal and monoclonal ant~bodles to CFTR showed that non-recomblnant T84 celis contain, in addHlon, a dHfuseiy miç~ratln~ 150kd (band C) version of CFTR.
It Is another aspect of the present Inventlon to study structure:function relatlonshlps In CFrR by constructln~ a sHe speclflc mutatlon whlch provldes for the deletlon of phenylaianine 508 (referred to as ~F508).
It Is yet another aspect to characterize variant Ci~rR protein forms associated wHh a number of less frequent CF assoclated mutatlons, as well as mutations In residues predlcted to pby an Important role In the functlon of Ci TR.
IG4-9.2 - ..:
- ~ .. -:
~ ~3 ^], ,~
It is still yet another aspect of the present inveniion to more fully describe the characteristics of CFTR associated with bands a, b and c.
It is yet still another asioect of the present invention to provide new diagnostic and therapeutic methods for CF which rely upon intracellular processing mechanlsm for CFTR and intracellular location of variousiy proce$ed CFTR.
SummarY of the Invention In accordance with the principles and aspects of the present invention there are provided recombinant DNA molecules encodin~ CFTR includin~ most preferred cDNA molecules whlch can be stably propagated in host E. coO cells and which canbe used to transform mammalian cells resuHin~ in expression of CFTR. These DNA
molecules are ideally maintained at low gene dosage in the host, thereby reducing the potential toxicity caused by inadvertentor inappropriate expression of the CFTR
cDNA. In addition, there are provlded recombinant cDNA molecules containin~ at least one intervening sequence w~lthin the Ci TR coding sequence. Such a sequence 2 o advantageousiy disrupts expression of protein from the Ci TR cDNA in E. con cells, but allows expression In mammalian celis slnce such cells are capable of removing the intervenin~ sequence from the initial CFTR RNA transcript. Also included are DNAsequences encodinçi Ci TR but containing one or more point mutations.
2 5 Preferred embodiments of the present invention inciude cDNA's codinçi for the entire Ci TR protein codinçi sequence of 4440 nucleotides and advantageousiy Include regulatory sequences from the fianking regions of the cDNA, such as the ribosome binding site located Immedlately upstream of the initlator methionine of the Ci TR open readinçj frame (Kozak, 1984; Kozak, 198b). These cDNA~s are ideally cloneci In plasmld vectors containinçi origins of replicatlon that allow maintenance of recombinant plasmids at low copy number in E. coll cells. These ori~lns of repllcation may be advantageously selected from those for the E. con plasmlds pMB1 ( 1~20 1G4-9.2 , ,. .
~ f - 1 copies per celi), pl5A (1~12 copies per cell) or pSC101 tapproximateiy 5 copies per celi) or other vectors which are maintained at low copy number (e.~. Iess than about 25) in E. co!i celis tSambrook et al., 1989).
Also described herein are ci~rR cDNAs containinS~ a synthetic intron of 83 base pairs between nucleotide positions 1716 and 1717 of the Ci TR cDNA sequence, which acts to stabil'lze the cDNA by disruptin~ the translational readin~ frame of the Ci-TR
protein such that neither full len~th protein nor extensive poiypeptide sequences can be synthesized In celis unable to splice mRNA. This allows repllcation in (but not Ci TR
expression) prokaryotic celis of the Ci-TR cDNA for subsequent transformation ofeukaryotic host ceils, most preferably mammallan cells, for subsequent Ci-TR
expression. Additional embodiments of the invention Include full len~th mutant Ci~TR
cDNAs which encode a protein from which amino-acid 508 has been deleted. Still other prefened embodiments include expression vectors for expression of said Ci-TR
cDNA's in bacterial, yeast, plant, insect and mammalian celis, and trans~enlc animais the Ci-TR proteins derived from these expression systems, pharmaceutical compositions comprisin~ such recombinantly prociuced Ci-TR proteins as well as assoclated dla~nostlc and therapeutic methods.
A most preferred embodiment includes mature Ci-TR protein, discovered to be assoclated with band c (described in detasil below) havin~i an apparent molecular welght o~ approximateiy 150kd and modified by complex-type N-linked ~iiycosylation at resldues 894 and/or 9~0. it has been unexpectedly discovered that mature Ci TR is lackln~ from recombinant celis encodin~i several mutant versions of the protein. Also 2 5 âescribeâ are new dla~jnostlc assays for detectlng Individuals sufferin~ from cystic flbrosls as well as therapeutlc methods for treatln~ such Indivldual based, in part, upon the mechanism of intracellular processlng of C--TR discovered In the present Inventlon.
1G4-9.2 . ~ . . . .:. :- -, - ' ' - ' /r~ t~
J L ~ ~
Brief Description of the Table and Drawin~s Further understandin~ of the invention may be had by reference to the tables and fl~ures wherein:
Table 1 shows the sequence of that portion of CFTi~ cDNA encodin~ the complete CFTR protein within plasmid pSC-CFTR2 Includin~ the amino acld sequenceof the CFTR open readin~ frame:
Table 2 shows CFTR mutants wherein the known association with CF (Y, yes or N, no), exon localization, domain location and presence (+) or absence t-) of bands A, B
and C of mutant cFrr~ species is shown. TM6. indicates transmembrane domain 6; NBD
nucleotide bindin~ domain; ECD, extracellular domaln and Term. termination at 21codons past residue 1337.
The convention for namin~ mutants is first the amino acid normally found at the particular residue, the residue number (Riordan et gL, 1989) and the amino acid to which the residue was converted. The sin~le letter amino acid code is used: D, aspartic acid; F, phenylalanine; G. ~iycine; I, isoleucine: K, Iysine; M. methlonine: N, aspara~ine; Q, ~lutamlne: R. ar~inine; S, serine; W, tryptophan. Thus 95~1D Is a mutant in which ~Iycine 551 is converted to aspartic acide:
Fi~ure 1 shows ali~nment of CFTR partial cDNA clones used in construction of cDNA contalnin~ complete codin~i sequence of the CFTR, only restriction sites relevant to the DNA constructions described below are shown;
Fi~ure 2 depicts plasmid construction of the CFTR cDNA clone pKK-CFTR1:
Fi~ure 3 deplcts plasmid construction of the CFTR cDNA clone pKK-CFTR2:
Fl~ure 4 depicts plasmid construction of the CFTR cDNA clone pSC-CFTR2:
Fl~ure 5 shows a plasmid map of the CFTR cDNA clone pæ-CFTR2:
., : .
~ 7 L,~ ~ ~
Fi~ure 6 shows the DNA sequence of synthetic DNAs used for insertion of an intron into the CFTR cDNA sequence, with the relevant restriction iendonuclease sites and nucleotide positions noted;
s Fi~ures 7A and 7B depict plasmid construction of the CFTR cDNA clone pKK-CFrR3:
Fl~ure 8 shows a plasmid map of the CFrR cDNA pKK-cFrR3 containing an intron between nucleotides 1716 and 1717;
Fi~ure 9 shows treatment of cFrR with ~Iycosidases;
Fl~ures 10A and 10B show an anaiysls of cFrR expressed from C0~7 transfected 15 cells;
Fl~ures 1 lA and 1 lB show pu~chase labellng of wild type and ~F508 mutant CFTR In C0~7 transfected celis;
Fl~ure 12 shows immunolocalization of wild type and ~F508 mutant CFTR; and C0~7 celis transfected wHh pMT-CFTR or pMT-CFrR-~ 8; and Fi~ure 13 shows an anaiysis of mutant forms of CFTR.
Detailed Descriotion and Best Mode Definltlons The term intron~ Identifies an Intervenin~ sequence wHhin a ~ene for the ~ene product that does not constHute proteln codlna sequences. In eukaryotic cells Introns are removed from the primary Ri~A transcript to produce the mature mRNA
The term splice- refers to the removal of an intron from the primary RNA
35 transcrlpt of a ~ene The term poiylinker refers a closely arran~ed series of synthetlc restriction enzyme cleava~e sHes wHhin a plasmld.
The term ~open readin~ frame- refers to a nucleotide sequence wnh the potentlal for encodin~ a protein S The term a~arose ~el purificanion refers to the separanion of DNA restriction fraçjments by electrophoresis through an agarose ~el followed by purificanion of the desired DNA fra~ments from the a~arose ~el as described below in ~eneral methods The term maintained- refers to the stable presence of a plasmid wnhin a transformed host cell wherein the plasmid is present as an aunonomously repncaning body or as an inte~rated portion of the host's ~enome.
The term cell cunure refers to the containment of ~rowin~ cells derived from either a municellular plant or animal which allows the celis to remain viable ounside of the original plant or animal The term host cell- refers to a microorganism including yeast, bacteria, insectand mammalian celis which can be ~rown in cell cunure and transfected or transformed wrth a plasmid or vector encodin~ a molecule having a Ci TR biological character'lstic.
The terms plasmid- and vector~ refer to an autonomous seif-replicating extra-chromosomal clrcuiar DNA and Includes both the expression and non-expression types When a recomblnant microor~anism or cell culture providin~ expre$ion of a molecule Is descrlbed as hostin~ an expresslon plasmid, the term expre$ion plasmid-Includes both extrachromosomal circular DNA and DNA that has been Incorporaned ~ -Into the host chromosome(s) The term promoter is a reglon of DNA invoived In bindlnçi RNA poiymerase to Initlate transcriptlon The temm DNA sequence refers to a sin~le- or double- stranded DNA molecule comprised of nucleotlde bases, adenosine (A), thymidine ~, cytosine ~C) and ~uanoslne (G) and further includes çjenomlc and complementary DNA(CDNA).
The term li~ate~ refers to the Joinin~ of DNA fra~ments via a covalent phosphodlester bond, whose formation is catalyzed for example, by the enzyme T4 DNA li~ase.
The term 'upstream identifies sequences proceedin~ in the opposHe direction from expre$ion: for example, the bacterial promoter is upstream from the transcription unH.
: ' The term 'restriction endonuclease-, aHernateiy referred to hereln as a restriction enzyme, refers to one of a class of enzymes whlch cleave double-stranded DNA
(dsDNA) at locatTons or sHes characteristic to the particular enzyme. For example the restrlctlon endonuciease Eco Ri cleaves dsDNA oniy at locations:
5'GAATTC3' to form 5~G and AATTC3~ fra~ments 3'CTTAAG5' 3'CTTAA G5' AHhou~h many such enzymes are known, the most preferred embodiments of the present Inventlon are primariiy concerned wHh only selected restriction enzymes 2 o havin~ specifled cnaracteristics.
All cited references are fuliy incorporated herein by reference, subsequent cHatlons of previousiy cHed references shall be by author oniy. Referenced citations, if not within the body of the text, may be found at the end hereof.
Wlthln Illustratlons of plasmld constructlons, oniy restriction endonuclease cleava~e sHes relevant to the partlcular constructlon beln~ depicted are shown.
Numberin~ of nuc~eotldes and amino aclds correspond to the published CFTR cDNA
sequence of Rlordan et al., complled from partlal CFTI? cDNA clones.
General Methocs Methods of DNA preparation, restrictlon enzyme cleava~e, restriction enzyme analysis, ~el electrophoresis, DNA preclpHation, DNA fra~ment li~ation, bacterial 35 transformatlon, bacterial colony selection and ~rowth are as detailed In Sambrook et aL DNA fra~ment iso atlon from a~arose ~eis was performed by crushin~ the agarose IG4-9.2 ~ i3 ~el slice containing the fra~ment of interest in 300 microrrters of phenol, freezing the phenol/~el slice m'xture at -70C for 5 minutes, centrifugini and separating the aqueous phase from the phend and extractlng the aqueous phase with chloroform.
The DNA fraS~ments were recovered from the aqueous phase by ethanol precipitation Methods of invitro transcription in a buffered medium and in vitro protein translation in rabbit reticulocyte Iysates were employed as detailed in the manufacturers instructlons (Strateç~ene and Prome~ ~a respectively). DNA sequencing was performed uslnç~ the San~er dideoxy method usin~ denatured double-stranded DNA (Sanger et al., Proc. Natl. Acad. Sci. 74, 5463 (1977)).
CFTR Partlal cDNA Source PartlalCiTRcDNAclonesT11,T16-1,T16-4.5andC1-1/5(Riordanetai.)were obtained from the American Type Cuiture Collectlon (Rockland, Maryland). A linear ali~nment of the CFTR cDNA portion of these clones is presented in Figure 1. Exons at the end of the Ind~vidual cDNA clones are indlcated by the numbers 1, 2, 7, 9, 12, 13 and 24. Aiso indlcated are the initiation codon of the CFTR protein coding sequence (ATG), the termination codon tTAG), as well as restfiction endonuclease sites withln the CFTR cDNA which were used in subsequent DNA manipulations.
Example 1- Generation of Full lenpth CFTR cDNAs Neariy all of the commonly used DNA clonin~ vectors are based on plasmids containin~ mod'lfied pMB1 replication origins and are present at up to 5ûO to 700 2 5 copies per cell (Sambrook et al.). The partial Ci TR cDNA clones Isolated by Riordan et al. were maintalned In such a plasmid. We postulated that an alternative theory to Intrlnsic clone instability to explain the apparent inabirriy to recover clones encoding full len~th CFrR protein usinç~ hi~h copy number plasmlds, was that it was not possible to clone lar~e segments of the CFTR cDNA at hi~h gene dosage in E. coll. Expression of the Ci ll? or poltions of the CFTR from regulatory sequences capable of directin~
transcriptlon and/or translation in the bacterial host cell might result In inviability of the host cell due to toxlcity of the transcript or of the full length CFTR proteln or fra~ments thereof, This inadvertent ~3ene expression could occur from either plasmid reçiulatory IG4-9.2 .
~ v 7 ~ 3 ~3 sequences or cryptic re~ulatory sequences w~lthin the recombinant cErR plasmid which are capable of functionin~ In E. col~. Toxic expresslon of the CETR codinç~
sequences would be ~reatiy compounded H a lar~ie number of coples of the Ci TR
cDNA were present in cells because a hi~h copy number plasmid was used. If ~he product was indeed toxic as postulated, the ~rowth of celis containin~ full len~th and correct sequence would be actively d'lsfavored. Bc~sed upon this novel hypothesis, the followln~3 procedures were undertaken.
With reference to i~içjure 2, partial CETR clone T16-4.5 was cleaved with restrlctlon enzymes ~ ! and i~st ! and the resuHlnS~ 3.9 kb restriction fragmentcontainln~ exons 11 throu~h most of exon 24 (Includin~ an uncharacterized 119 bptnsertlon reported by Riordan et al. between nucleotldes 1716 and 1717), was isolated by a~arose ~el puriflcation and li~ated between the Sph ! and Pst ! sites of the pMB1 based vector pKK223-3 (Brosius and Hoiy, i~roc. Natl. Acad. Scl. 81, ~929 (1984)). It was hopeâ that the pMB1 oriç~in contained wnhln this plasmid would allow it and plasmids constructed from it to replicate at 15-2û coples per host E. con cell (Sambrook et al.).
The resunant plasm'ld clone was called pKK~.5.
Partlal Ci rR cione T11 was cleaved with Eco Rl and Hlnc n and the 1.9 kb band 2 0 encodln~ the flrst 1786 nucleotides of the Ci TR cDNA plus an additlonal 100 bp of DNA at the 5' end was isolated by a~arose ~el puriflcatlon. This restrictlon fragment was Inserted ioetween the Eco Rl site and Sma ! restrlctlon site of the plasmid pBluescript SK- (Strate~ene, catalo~ue number 212200. such that the ci~rR sequences were now flanked on the upstream (5') slde by a Sal ! s'lte from the clonln~ vector. This clone, desl~nated T11-R. was cleaved wHh Sal ! and SDh ! and the resuHant 1.8 kbband isolated by açiarose ~el puriflcatlon, i~asmld pKK-4,5 was cleaved with Sal ! and Soh ! and the lar~e fra~ment was Isolated by a~iarose ~7el puriflcatlon, The purified T11-R fra~iment and pKK-4.5 fra~ments were ll~ated to construct pKK-Ci TR1. pKK-Ci ~R 1 contalns exons 1 throu~h 24 of the Ci~R cDNA. It was discovered that this plasmld Is stably ma~ntalned in E. coll celis and confers no measurably disadvanta~eous ~rowth characteristlcs upon host celis.
IG4-9.2 pKK-CFTR1 contains, between nucleotides 1716 and 1717, the 119 bp insert DNA
derived from partial cDNA clone T16 4.5 described above. In additlon, subsequentsequence analysis of pKK-Ci TR1 revealed unreported dmerences In the codin~
sequence between that portion of Ci-~?1 derived from partiai cDi~A clone T11 and the publ'lshed CFTR cDNA sequence. These undeslred d'lfferences included a 1 base-pair deletion at position 995 and a C to T transHion at position 1507.
To complete construction of an intact correct CFTR codin~ sequence wlthout mutations or insertlons and with reference to the construction scheme shown in Fi~ure 3, pKK-CFTR1 was cleaved with Xba ! and Hpa ! and dephosphorylated with caif intestinal alkaline phosphatase. In addition, to reduce the likelihood of recoverin~ the ori~inal clone, the small unwanted Xba !/Hpa ! restriction fra~ment from pKK-CFTR1 was dl~aested with Sph 1. T1~1 was cleaved with Xba ! and Acc ! and the 1.15 kb fra~ment isolated by a~arose 5~el purification. T1~4.5 was cleaved with Acc ! and Hpa ! and the 0.~5 kb band was also isolated by aqarose qel pur'lfication. me two a~arose ~el purifled restriction fra~ments and the dephosphorylated pKK-CFTR1 were li~ated to produce pKK-CFTR2. Aiternatively, pKK-CFTR2 could have been constructed usin~ correspondin~ restriction fra~jments from the partial CFTR cDNA clone C1-1/5.
2 o pKK-CFTR2 contains the unlnterrupted Ci TR protein codinq sequence and conferred slow ~rowth upon E. coll host cells in which H was inserted, whereas pKK-Ci~TR1 did not.
The ori~in of replication of pKK-Ci TR2 is derived from pMB1 and confers a plasmid copy number of 1~20 coples per host cell.
ExamPle 2 - ImprovinQ Host Cell ViabllHy An addHlonal enhancement of host cell viability was accomplished by a further reduction in the copy number of Ci~rR cDNA per host cell. This was achieved by transferrin~j the Ci TR cDNA into the plasmid vector, pSC-3Z. pSC-3Z was constructed usln~ the pSC101 replication ori~in of the low copy number plasmld pLG338 (Stoker et al., Gene 18, 335 (1982 and the ampicillin resistance ~ene and poiylinker of pGEM-3Z
(available from i~rorne~a). pLG338 was cleaved with SPh ! and i'vu ll and the 2.8 kb fra~ment contalnin~ the repllcation ori~in isolated by a~arose ~el puriflcation. pGEi~
3Z was cleaved wHh Aiw Nl, the resuHant restriction fra~ment ends treated wHh T4 Di~A
~5 poiymerase and deoxynucleotide trlphosphates, cleaved wHh Sph I and the 1.9 kb IG4-9.2 . , ~.
band containing the ampicDlin resistance gene and the poiyiinker was isolated byagarose çiel purification. Thle pLG338 and pGE~3Z fragments were li~ated together to produce the low copy nurnber cloning vector pSC-3Z. pSC-3Z and other plasmidscontaininç; pSC101 origins of replication are maintalned at approximateiy five copies 5 per cell (Sambrook et aL).
With addit'lonal reference to Fi~ure 4, pKK-CFTR2 was cleaved with Eco RV, Pst !and Sai ! and then passed over a Sephacryi S400 spun column (availabel from Pharmacia) accordin~ to thle manufacturer~s proceciure in order to remove the Sal !
to Eco RV restrict~lon fragment which was retalned within the column. pSC-3Z wasdiçjested with Sma ! and Pst ! and also passed over a Sephacryl S400 spun column to remove the small Sma !/Pst ! restrictlon fragment whlch was retained wHhin the column. The colurnn eluted fractions from the pKK-Ci-Ti?2 diçlest and the pSC-3Z diges, were mlxed and iigated to produce pSC-CFTR2. A map of this plasmld is presented In Figure 5. Host celis containin~ CFTR cDNAs at thls and slmllar ~ene dosages çjrow well and have stabiy rnalntained the recomblnant plasmld with the full length Ci TR coding sequence. In addition, this plasmid contains a bacteriopha~e T7 RNA poiymerase promoter adJacent to the Ci-TR coding sequence and Is therefore convenlent for in vitro transcription/transiation of the CFTR proteln. The nucleot~lde sequence of Ci-TR
2 o codinçi region from pSC-CFTR2 plasmid Is presented in Table 1. Significantly, this sequence differs from the previousiy published (Riordan et al.) CFTR sequence atpositlon 1991, where there is C in place of the reported A. E. coli host celis containing pSC-CFTR2, Internally ~dentified with the number pSC-CFTR2/AG1, have i~een deposited at the American Type Culture Collection and given the accession number:
Example 3 - Aiternate Method for Improvin~ Host Cell Viabllity A second method for enhanclnçi host cell vlabllity comprises disruptlon of the 3 o CFTR protein coding sequence. For thls purpose, a synthetlc Intron was desl~ned for Insertlon between nucleotldes 1716 and 1717 o~the CFTR cD~iA. This Intron is especbllv advar~tageous because of its easiiy manageable size. Furthermore, it Is deslgned to be emcientiy spiiced from CFTR primary RNA transcripts when expressed in eukaryotlc celis. i-oursynthetlc ollgonucleotldes were synthesized (1195RG,1196RG, IG4-9.2 2 ~
1197RG and 1198RG) collectively extendin~ from the Soh ! cleava~e site at posltion 1700 to the Hlnc 11 cieavaçie sHe at posHion 1785 and Includinç~ the additlonol 83 nucleotldes between 1716 and 1717 (see Fiç~ure 6). These oliçlonucleotides were phosphorylated with T4 polynucleotide kinase as described by Sambrook et al. mixed to~ether heated to 95 C for 5 minutes in the same buffer used durinç~
phosphorylation and allowed to cool to room temperature over several hours to allow annealin~ of the sinçlie stranded oRç10nucleotides. To Insert the synthetlc intron into the CFTR codinç~ sequence and wHh reference to F~ures 7A and 7B a subclone of plasmid T11 was made by cleavinçl the Sal ! site In the poiyiinker repairinçl the recessed ends of the cleaved DNA with deoxynucleotlde triphosphates and the large fraç~ment of DNA Poiymerase I and reliçlatln~ the DNA. This plasmld was then di~ested wHh Eco RV and Nru ! and reli~ated. The resuitlnçi plasmid T16-~ 5 extended from the Nru ! site at position 49û of the CFTR cDNA to the 3 end of cione T16 and contained sin~le sHes for SPh ! and Hinc 11 at posHions correspondinçl to nucleotides 1700 and 1785 of the CFTR cDNA. T16-~ 5 piasmid was cleaved with Sph ! and Hinc 11 and the lar~e fraçiment was isolated by aç~arose ç~el purification. The annealed synthetic oliç~onucleotides were liçicsted into this vector fragment to çlenerate T1~intron.
T16-intron was then di~ested with Eco Rl and Sma ! and the iarge fragment was isolated by açlarose ~el purification. T16-4.5 was d~gested with Eco Rl and Sca ! and the 790 bp fraçlment was also isolated by açlarose çlel purification. The purlfied T16-lntron and T16-4.5 fra~ments were li~ated to produce T16-intron-2. T16-intron-2 contalns CFTR cDNA sequences extendinçl from the Nru ! site at position 490 to the Sca ! sHe at posHion 2818 and Includes the unique H~a ! sHe at positlon 2463 which is not present In T16-1 or T1~intron-1.
T16-lntron-2 was then cleaved wHh Xba ! and H~a ! and the 1800 bp fraçlment was Isolated by a~arose ~el purification. pKK-CFTR1 was diçlested with Xba ! and Hpa ! and the lar~e fra~ment was also isolated by aS~arose ~el purification and ll~ated with the fra~ment derived from T16-intron-2 to yield pKK-CFTR3 shown in Fiçjure 8. The CFTR
cDNA withln pKK-CFTR3 is ldentical to that wHhin pSC-CFTR2 and pKK-CFTR2 except for IG4-9.2 J~ 't the insertion of the i33 bp intron between nucleotldes 1716 and 1717. The Insertion of ` ' this intron resuited in improved çjrowth characteristlcs for cells harborinçj pKK-CFTR3 relatlve to cells containin~ the unmodified Ci TR cDNA in pKK-Ci-~R2.
Example 4 - In vitro Transcription/Transiation In addition to seciuence anaiysis, the Inteç;rity of the Ci TR cDNA open readin~frame was verified by in vHro transcription/translatlon. Thls method aiso provided the InHlal CFTR protein for Ident'lfication purposes. 5 mlcro~jrams of pSC-CFTR2 plasmld DNA were llnearized with Sal! and used to dlrect the synthesis of CFTR RNA transcrlpts wHh T7 RNA polymerase as described by the suppller (Strata~ene). Thls transcript was extracted with phenol and chloroform and preclpitated with ethanol. The transcript was resuspended in 25 microliters of water and varyinçl amounts were added to a retlculocyte Iysate In vitro translation system (from Prome~a). The reactions were performed as described by the supplier In the presence of canine pancreatic microsomal membranes (from Promeçla), uslnçl 35S-methlonine to label newly synthesized proteins. In vitro translation products were analysed by discontinuous poiyacrylamide Çiel electrophoresis in the presence of 0.1% SDSwith 8% separatinç
çleis (Laemmll, 1970). Before electrophoresis, the In vitro translation reactions were denatured with 3% Si~S, 8 M urea and 5% 2-mercaptoethanol In 0.65 M Tric-HCI, pH 6.8.
Followlnçi electrophoresis, the çleis were fixed in methanol:acetlc acld:water (30:10:60), rinsed wHh water and Impreçlnated wHh 1 M sodlum salicylate. 36s labelled proteins were detected by fluoro~raph. A band of approxlmateiy 180 Kd was detected, consister~t with translation of the full lençlth CFTR Insert.
Example 5 - Ellmlnation of crvPtlc Repulatorv Sl~nais Analysis of the of the DNA sequence of the CFTR has revealed the presence of a potentlal Ej. coll RNA poiymerase promoter between nucleotldes 748 and 778 whlchconforms well to the derived consensus sequence for E. cOn promoters (Reznlkoff and McClure, Maxlmizin~ Gene Expresslon, 1, Butterworth i'ubllshers, Stoneham, MA). If th~3 sequence functlons as a promoter functlons In E. coll, H could dlrect synthesls of potentlaliy toxlc pariial CFTR polypeptldes. Thus, an addHional advantaçleous procedure for maintalnln~ plasmlds contalnln~ CFTR cDNAs In E. cOn would be to aiter IG4-9.2 2 ~ ~ 7 ~
the sequence of this potential promoter such that H wlll not function in E. coll. This m~
be accomplished without alterinçj the amino acid sequence encoded by the Ci TR
cDNA. More, specifically, plasmids containin~ complete or partial Ci-TR cDNA~s would be aitered by slte-directed muta~enesis usin~i synthetlc oii~onucleotides (Zoller and s Smith, Methods Enzymol. 100, 468, 1983). Specifically, aiterin~ the nucleotide sequence at positlon 748 from a T to C and at position 774 from an A to a G effective~
ellmlnates the actlvity of this promoter sequence without aiterin~ the amino acid codin~ potential of the Ci~rR open readin~ frame. Other potentTal re~ulatory si~nais within the Ci TR cDNA for transcription and transiation could aiso be advantageously aitered and/or deleted by the same method.
Example 6 - Clonln~ of Ci-TR in aiternate host systems AHhou~h the CFTR cDNA dlsplays apparent toxicity in E. coli celis, other types of host celis may not be affected in this way. Aiternative host systems in which the entire Ci TR cDNA protein encodin~ re~jion may be malntalned and/or expressed include other bacterial species and yeast. It is not possible a priori to predict which cells mi~;int be resistant and whlch mi~jht not. Screenin~ a number of different host/vector comblnations Is necessary to find a suitable host tolerant of expression of the full length 2 o proteln or potentlally toxic fra~ments thereof.
Example 7 - i~roduction of Ci-TR mutants and relevant plasmid constructions Mutatlons were Introduced Into Ci-TR at resldues known to be aitered In CF
chromosomes (~F508, ~1507, R334W, S5491, G551D) and in residues belleved to play an Important rols In the function of cFri? (K464M, F508R, N894,900Q, K1250M). CFTR
encoded by these mutants was examined in COS-7 celis transfected wHh cDNA
plasmlds havln~ the aforementloned aiteratlons. Remarkably, H was surprisingly discovered that mature, fuliy ~iycosyiated CFTR was absent from celis containin~~F508, ~1507, K4~4M, F508R and S5491 cDNA plasmlds. Instead, an unstable, Incompietejy ~jiycosyated verslon of the proteln was detected wHh an apparent molecular wel~jht of 1 35kd. Surprisin~jiy, the immature, mutant versions of cFrR appear to be reco~jnked as abnormal by a component of the post-translatlonal Intracellular IG4-9.2 transport machiner~, and remain incompletely processed in the endoplasmlc reticulum where they are subsequentiy degraded. Slnce mutations wHh this phenotype represent at least 70% of known CF chromosomes, we have discovered that the primary couse of cystic flbrosis Is the absence of mature CFTR at the correct cellular location, see aiso Figures 10 and 12. As a resuH of this surprisin1 resuit, thls Inventlon provldes new approaches to the dia~nosis and treatment of CF.
Recomblnant DNA manipulatlons were per-ormed accordini to standard methods (Sambrooi~ et aL,1989). Olijonucleotlde-directed mutagenesis of the cFrR
o cDNA was performed os described by Kunkel (1985). A piasmid vector for CFrR
expression In mammalian cells was constructed by placlng CFrR cDNA sequences from the Ava ! sHe at posHlon 122 In the cDNA sequence to the Sac ! sHe at posHlon 4620 Into the unlque i~HI sHe of the expression vector pSC-Ci-V1 uslng syntheticadaptor sequences Tile resulting plasmid was called piur-cFrR. In piVir-CFTR, expression of cFrR is controlled by the flanking mouse metaliothlonein-l promoter and SY~iO eariy poiyadenylation siç;nal. rhe vector also contains an origin of replication from pSC101 (Cohen,1973) for replication In E. coll, the i~lactamase ~ene and anSV40 orlgin of repikation. For convenlent sHe-d~rected mutagenesis of CFn?, the cryptlc bacterial promoter wHhin the CFTR cDNA of plasmid pTM-CFTi?-3 (Gregory et aL,1990) was flrst inactivated by changing the T resldue at nucleot~de 936 to a C such that plasmlds contaWn~ CFT~? sequences could be maintained at hi1h copy number wHhout corresponding change in amino acld sequence. The cFn? cD~iA was then Inserted between the Aoa ! and Sac ! sltes of the hljh copy number vector pTivi-1 (avallable from T. Mizukaml, O. Elroy-Steln and B. Moss, National Institutes of HeaHh) 2 5 uslng a 5 flanklng Apa ! sHe common to pTM-CFn?-3 and pTM-1, and the Sac ! sHe at posHlon 4620 In tne CFTR cDNA. This plasmld, pTM-CFTR~, was used for all subsequent mutagenesis of the CFTi? sequence. For expresslon In COS-7 celis, CFri? cDNA mutants constructed In pTi~CrTR-4 were dlgested wHh Xba ! and BstX ! and the 3.5 kb CFTRcDNA fragment wc~s purified and placed between the unlque Xba ! and BstX ! sHes withln the CFTI? ci~A portlon of pMT-CFTR. Translent expression of cFrR In C0~7 celis was performed essentialiy as described by Sambrook et aL,1989.
IG4-9.2 7 i~
Example 8 - Production of Ci TR and Protein TheraPy Protein therapy may be accompllshed by usln~ CFTR proteln produced by host celis transformed or transfected wHh the CFTR cDNA of the present Invention to s correct the CF defect directly by Introducin~i the proteln into the membrane of cells lackln~ functional CFTR protein. This therapeutlc approach au~jments the defective proteln by addHion of the wild-type molecule. The full len~th cDNA disclosed here can readliy be used via conventional technlques to produce vectors for expression of the CFTR proteln In a variety of well known host systems. Protein or membrane fra5~ments purifled or derived from these celis can be formulated for treatment of cystic flbrosis.
Recomblnant cFrR can be made usin~ techniques such as those reported by Numa (Harvey Lectures 83,121 (1989) and references cHed thereln) forthe synthesis ~f other membrane proteins under the directlon of transfected cDNAs. It will be Important to realr~e that toxicHy can result in mammalian celis from over expresslon of membrane protelns (Belsham et ai., Eur. J. Biochem. 156,413 (1986. Fortunately, to clrcumvent the potentlal toxicity of the protein product, vectors wHh inducible promoters (Klessl~ et al., Mol. Cell. Biol. 4,1354 (1984)) cna be advanta~eousiy used.
For exampie, for constitutive expresslon in mammallan celis, the full len~th ci~rR
cDNA clone is constructed so that it contains Xho ! sHes Immedlately 5~ to the Initiator methlonlne ATG and 3' to the terminator TAG. These sHes are unique since there are no Xho ! SneS In the CFrQ cDNA sequence. This facilitates Incorooratlon of the DNA
sequence encodin~ CFTQ into the expression vectors of the types described below.
Those skllled in the an will reco~nize that many possible cell/vector systems have been used successfully for the hl~h level expresslon of recomblnant proteins. Several sultable systems are descrlbed below. i30vine Papilloma Virus (ilPV) based vectors (Hamer and Wallin~, J. Mol. 8~ Appl. Gen. 1,273 (1982 can oe used to transform mouse C127 celis. C127 celis comprise an adenocarclnoma cell line isolated from a mamma~y tumor of an R111 mouse (ATCC: CQL 1~10. Foliowin~ the procedures of Hslun~ et al. (J. iviol. & Appl. Gen. 2,497 (1984 and Qeddy et al., (DNA k 4~1 (1987.
IG4-9.2 : :
. . :,~
, . . ..
S~ 7 the i~PV vector can be constructed in such a way as to expre$ recombinant CFTr?
proteln under controi of the mouse metallothionlne promoter and poiyadenylatlon sequences. Once a construct containing the CFTR cDNA is made, it is then advanta~eously transfected into the C ~ 27 celis usln~ standard calclum phosphate precipitatlon methods (Graham and Van der Eb, Virolo~y 52,456 (1973)). The transformed celis can then be selected by foc~ forrnation. A similar vector, In which the ~ene for neomycin resistance (Southern and i3er~, J. Mol. & Appl. GenM,327 (1982)) has been Inserted into the unique Sal 1 site, may advanta~eousiY also besuper-transfected into the same cells and celis Incorporatin~ such vectors suHably 10 selected with the antlbiotic G418. This method convenientiy decreases the time necessary to select for desired cell lines expressin~ the transfected ~ene product.
Another expression system empioys vectors In which the cDNA is under control of the metallothionine ~ene promoter and the SV40 eariy polyadenylation si~nal. In 15 addltlon, the mouse dihydrofolate reductase (DHi-R) cDNA (Nunberçl et al., Cell 19, 355 (1980 Is under controi of the SV40 eariy promoter and poiyadenylation si~nal.
This vector Is then idealy transfected into Chinese Hamster Ovary (CHO) cells (ATCC:
CCL 61) that are deficient in DHFR (Uriaub and Chasin, i~roc. Natl. Acad. Sci. Z,4216 (1980)). Transformed cells can be selected and the Ci T ? contalnin~ vector 20 sequences amplifieci by cuHurin~ the celis in medla containln~ the dru~ methotrexate Yet another example of an inducible expresslon system invoives the use of vectors based upon the commerclaliy available plasmid, pMAivineo (Clontech).
pMAMneo contains a mouse mammary tumor virus promoter for expression of cloned 25 ~enes. This promoter can be Induced by treatin~ transfected celis with çjlucocorticolds, sucin as dexamethasone, resuH~nç; in elevated expression of the cloned ~ene. The ~a+/H+ antiporter is a membrane iorotein that Is structurally very simllar to the Ci TR ar~d has been successfuliy expressed with the pMAMneo vector (Sardet et aL, Cell 56,271 (1989)). Vectors based on pMAMneo, but containin~ low3 0 COpy number E. coii ori~ins of replication, could be used for Induclble expresslon of CFTR In elther C127 celis, CHO or other mammalian cells as described above.
Similariy, many suitable expression vector/host systems have been described for the expression of mammallan proteins in bacteria, fun~l, insect and plant cells and in the milk of trans~enic animais. One skilled in the art can modify these expression s systems for the production of CFTR. For exampie, low copy number CFTR vectors, based upon the Invention described hereln, couid be used to direct synthesis of CFTR
protein in E. coD. To avoid toxicity due to expression of CFTR R~iA or protein, the CFTR
cD~iA must be under the transcriptional control of a re~ulatable promoter. As anexample of one such Inducible expresston system. the T7 RNA polymerase promoter o within pSC-CFTR2 could be used to Induce transcription of CFTR sequences in E. coil as described byStudierand Moffat (J. Mol. i3ioi.189,113 (1986). In orderto maximizeleveis of CFTR protein expression after transcriptional Induction, it would be necessary to introduce an E. coli ribosome bindin~a site (Shine and Dalgarno, Nature 254,43 (1975)) upstream of the CFTR initiator methlonine. Prokaryotic orSIanisms other than E.
coll could aiso be used for expression of Ci TR protein. For example, a membrane-bound phosphotriesterase has been successfully produced In strePtomyces lividans by Steien et ai. tB~otechnolo~y 7,65 (1989)).
Owin~ to the nature of CFTR ~Icosylation, the most preferred expression systems Will utilize mammallan celis. Translent expression of C~TR can be accomplished using C0~7 celis as previouisiy described In Exampie 7 and In subsequent examples.
Forei~n proteins have been expre$ed usin~ a variety of vectors in many different fun~l. For example, van den Ber~ et aL (Blotechnolo~y 8,135 (1990)) have produced prochymosln in Kluvveromvces lactis, Loison et al. (Blotechnology 6,72 (1988 produced hirudin In Saccharomvces cerevisiae. and Cre~3 et aL
(Blotechnolo~y 5,479 (1987 have produced hepatHis B surface antl~en in Plchla pastoris.
For Insect celis, the ;3-adrener~lc receptor, a membrane proteln, has been expressed usln~ a baculovirus expresslon vector (Geor~e et al., Biochem. Biophys. Res.
Comm. 163,1265 (1989. CFTi? could be produced in insect cells by obvious modH`icatlon of this system.
IG4-9.2 - . .
. . .. . . . - .. ~ .
-cFrR could be expressed in plants by modification of the techn~iues of Hiatt e~aL (Nature 342, 76 tl989)) whlch have demonstrated the production of the Immunoç11obulin heavy and liçiht chalns In tobacco and other plants.
Techniques for the production of forelçln proteins in the milk of trans~enic anlmals have also been described in EPA 0264,166, fuliy incorporated herein. These technlques can readiiy be modified for productlon of CFTR in the milk of mammals.
Slmilariy, the Invention described hereln enables the use of technlques known to thosa skllled in the art forthe productlon of a transç1enlc anlmal mocel for cystlc fibrosis.
Such a CF animal model could be advanta~eousiy employed to screen for suHable pharmacoloçllcal therapeutlc açjents as later described.
Example 9 - Characterization of the cFrR Protein A. Isolation of cFrR.
cFrR is a membrane protein havinç~ an amino acld sequence which contains reçilons with extensive hydrophoblc character. In order to purify CFTl? as a functlona~
2 o proteln H wlll be important to accomplish the solubilizatlon of the CFTiR from its native membrane such as throu~h the use of deter~ents.
CondHions for the solubllizatlon of CFTR from Hs natural lipid environment can i~e advantaçleousiy determined usln~ whole celis, or membrane preparatlons prepared from cells whlch express C~ll?. As wlll be readiiy understood, initlal solubllkatlon experlments wlll Invoive screenlnç~ a variety of deterçients at varyinç~ concentratlons 'n order to flnd condmons that preferabiy achleve optlmal solubilizat~on of the Ci rR.
Briefly, packed memiorane pellets are resuspended In deterçlent solution, ~entlyhomo~enized, and the insoluble material removed by centrifuçiation at 1OO,a~i for one hour. The de~ree of solubllkatlon achleved is Ideally monitored Immunolo~lcaliy.
Potentlal deter~ients Include, but are not llmHed to, CHAi S (3-(3-cholamidopropyi)aiimethyiammonlo)-l-protanesuifonate) (Borsotto M., et a'., J. Blol.
Chem, 260, 14255 (1985)), Hamada and Tsuro, J, Blol. Chem. 263 1454 (lq88)), n-octyi ~lucoslde (Landry et ari., Sclence 24~i, 14~9 (1989: lubrol (Smlçiel, J. Biol. Chem. 261, IG4-9.2 , i3 (1989). Brlefly, Immunopreclpitates were incubated with 20 n~ of protein kinase A
(Si~ma) and 10 jlCI of (~32P)ATP in 50 j~li of kinase buffer (50 mM Tri~HCI, pH 7.5,10 mU
MçjCI2 and 100 i~lg/ml bovine serum albumln) at 30C for 60 minutes. The reaction was stopped by the aWitlon of 0.5 ml RIPA buffer (50 mM Tris-HCI, pH 7.5, 150 mM NaCI,1%
Triton X-100,1% sodium deoxycholate and 0.1% sodlum dodecyl sulphate). The procedure for Cleveland diçjestion was performed as described by Cleveland et al.
(1977) with modifications (Chen~ et _.,1988).
C. Di~estion with ~IYcosidases.
The ~Iycosidases N-GLYCANASE(R) enzyme, O-GLYCANASE(R) enzyme, endoçjiycosidase H and endoçjlycosidase F were obtalned from Genzyme Corporation. CondHions for dlçjestlon with the respective enzymes were as specified by the manufacturer except Incubations were performed at 37C for 4 h only. All dl~estions were performed on CFTR which had been purified by immunoprecipitationand separation on poiyacrylamide ~eis (see Example 10). CFTi? bands B and C wereeluted from the ~aeis by maceratlon of the ~el p~eces in extraction buffer (50 mM
ammonium blcarioonate,0.1% SDS and 0.2% 3-mercaptoethanoi). Referrin~ to Fiçjure9, bands B and C were Immunopreclpitated from T84 celis and phosphoryiated in vHro 2 o usinçi proteln klnase A and (~32P)ATP. The CFrR proteins were extracted from the SDS-poiyacryiamlde ~eis, subJected to no treatment (lanes 1,3,5 and 7) or were Incubated with i~GLYCANASE(R) enzyme (lanes 2 and 4), endoçilycosidase F (lane 6) or endo~ilycosidase H (lane 8). Samples were separated by electrophoresis and analysed by autoracitoçiraphy. Exposure was for 24 h.
D. Pulse-chase studles.
Six 90 mm dishes of COS-7 cells were transfected wHh eHher pMT-CFTR or pMT-CFTR-~F508. To avoici dish to dish variatlon In transfection efficiency, at 12 h post-transfectlon, the ceils were harvested by trypsinizatlon and re-distributed amonçi six 90 mm dishes. Followir~ 18 h of Incubatlon, the celis were washed twice wHh DME media (lacklnçi methlonlne) and starved for 30 mlnutes at 37C. (35S)methlonlne (250 ilCi/ml) was then added to each dlsh and the plates iaoeled for 15 mlnutes at 37C. At the IG4-9.2 , , ~ ' ' '' , ' -, . .
1976 (1986)); A~new et al., BBRC 92,860 (1980)) Triton X-100 (Hartshonne and Catterall, J. Biol. Chem. 259,1667 (1984)); and Triton X-114 (Bordier, J Biol Chem 256,1604 (1981)).
The initial detergent solubilized CFTR solution can also be diluted into an appropriate concentration of detergent or deter~ent/lipid (Agnew and Raftery, Biochemistry 18, 1912 (1979)) to achieve stabilization of the CFTR. Compounds known to stabilize proper folding of membrane proteins, sometimes refened to as ozmoiytes, can also be used. Such stabiiization a~ents include poiyols such as giycerol, sugars and amino acids (Ambudkar and Maioney, J. Biol. Chem. 261,10079(1986)). In addition, protease inhibitors a~ainst the four major classes of proteases are advantageousiy present throughout these procedures (Hartshorne and Catterail, J. Biol. Chem. 259,1667 (1984)) and would include, for example, phenylmethyisuifonyl fiuoride for serineproteases; iodoacetamide forthiol proteases; 1,10-phenanthroiine for metailoproteases; and pepstatin A for proteases wrth activated carboxylic acid 5~roups. ideally, studies should be carried out in which the concentrations and reiative proportions of deter~jent, lipid and ozmolyte are varied together with other buffer condnions in order to identify optimai conditions to preserve and stabilize the CFTR.
For example, A~new and Raftery varied the ratio of various detergents and lipids and determined that a 7 to 1 ratio of lubrol to phosphatidylcholine stabilrzed the solubilrzed volta~e sensitive sodium channel for further purification. Simiiariy, Hartshorne and 2 o Catterail found that the presence of 0.25% e~ phosphatidylcholine produced a more stable preparatlon anà an Increased recovery during purification of the sodium channel solubillzed with Triton X- 100. To determine the functionai integrity of the solubiirzed protein may require reconstitution of the protein Into liposomes using the procedure of Example 11, foilowed by introduction into cells and testing using the ion effiux assays of Example 14.
B, ImmunoprecipHations and protein phosphorvlation usina protein kinase A.
The procedures employed for isotopic labelin~ of celis, preparation of ceii 30 Iysates, immunopreclpitation of proteins and SDS-poiyacrylamide gel electrophoresis were as described by Chençl et al.,1988 and Gre~ory et aL,1990. CFT ? was phosphorylated in vHro wHh protein kinase A essentially as described by Kawata et al.
iG4-9.2 ~, f ~) 6 " ~ i è ~
end of the 15 minutes, the celis were washed twice wHh çjrowth medla, maintained in growth media and then chased for various times up to 24 h. Referring to Fiçjure 11 A, COS-7 celis were mock transfected (lane 1) or transfected with piviT-Ci-TR (lane 2), pMT-CFTR-~F508 (lane 3) and pMT-CFTR-Tth 1111 (iane 4). 48 h post-transfection, the s celis were labeled for 12 h wHh ~35S)methionlne. CFTR from these iysates were immunoprecipHated wHh the monoclonal antibody mAb 13-1 (see Exampie 11) and then analyzed on a SDS-polyacrylamide gel. The çjel was fluorographed and exposed for 4 h. In Fiçjure 11 B COS-7 ceiis were either transfected w'lth pMT-CFTR (lanes 1 -6) or pMT-CFTR-~F508 (lanes 7-12). At 48 h post-transfection, the celis were labeled for 15 minutes wHh t35S)methionine. After being iabeled, the celis were eHher harvestedimmediateiy or rinsed several times wHh labeling media, transferred to standard growth media and then harvested at various t~mes thereafter. The Iysates prepared were immunoprecipHated with mAb 13-1 and analyzed on a SDS-polyacrylamide gel.
The fluoroçjraph çiel was exposed for ~ h.
E. Immunofluorescence microscoPy~
Indirect Immunofluorescence was performed essentially as described by Kalderon et al. (1985). COS-7 cells whlch had been transfected wHh CFTR-containing cDNAs (see Example 7) were transferred onto ~iass coverslips at 12 h. Following a further 18 h Incubation at 37C, the cells were fixed In 3.7% formaldehyde in phosphate buffered sallne (30 minutes at room temperature), permeabilized wHh 1 %
Nonldet P40 (15 mlnutes at room temperature) and Incuioated wHh the monoclonal antibody mAb 13-1 (see Example 11) followed by FiTC-conJugated goat anti-mouse IgG (Cappel Labs.). The cover sllps were mounted uslng 50% glycerol in phosphatebuffered saline and viewed usinçi a Zeiss Axioplan mlcroscope. With reference toFlgure 12,48 hours after transfectlon, the cells were flxed and stalned using the monoclonal antibody mAb 13- 1 tExample 11) or 423 (specHflc for S\/40 Larçj~T antigen) as flrst ant~body. The second antlbody was fluorescein-conjuçiated goat anti-mouse 3 16iG. The localkation of the va~ous CFTI~ protelns were visualized by Immunofluorescence mlcroscopy. Mlcrographs show (A) shows nuclear stainiin~j of SV40 Large-T antlgen us~ng the monoclonal antibody 423 (Harlow et al.,1981): (B)shows pMT-CFTR Incubated wHh mAb 13-1 In the presence of excess fuslon proteln: (C) Shows pMT-CFm-~F508 Incubated with mAb 13-1 and (D) pMT-CFm Incubated wHh mAb 13-1.
IG4-9.2 2~3~ 3 Example 10- Puriflcation of the CFTR Protein Utilkin~ the soiubil'zed CFTR protein from Example 9, one may purify the CFR
utilizin~ puriflcatlon procedures which have been employed previously with slmilar membrane proteins. Aithouç1h proteins wHh muitiple membrane spanning domains have been purifled using conventional technlques (Catterall, Science 242 50 (1988)), the ~eneration of speciflc antibodies has allowed other investiç1ators to develop rapid and slmple purHflcation schemes for P-çjlycoprotein (Hamada and Tsuro, J. Biol. Chem.
263 1454 (1988)), and sodium channels (Casadel et al., J. Biol. Chem. 261 4318 (1986):
Nakayama et al., Proc. Natl. Acad. Sci. 79 7575 (1982)). Thus, the production of CFTR
speciflc antibodies (see Example 11) could faciiHate the purification of the CFTR
molecule and allow Hs puriflcation away from the relatlvely hi~h level of contaminants expected in the startin~ solubilized preparation.
For example, antibodies produced against an extracellular or other domain of the CFTR could be screened to select therefrom an antibody havin~ a suitably hi~h bindin~ coefflcient appropriate for use in the puriflcation scheme. The selectedantibody is ideally immobiiized on a variety of commercialiy available resins including 2 o CNBr activated Sepharose, Af~Gel 10, Reacti-Gel CDI and Amino-Unk resins and tested for immobilized antibody capacity. Optimal conditions for bindin~ CFTR to the column, washin~ the column to remove contamlnants, and elutinçl the purHfled prote'r can then be determined usin~ conventional parameters as the startin~ point and testinç~ the effect of varying the parameters. it wiil be reco~nked that effective wash and elution condHions will si~iniflcantly impact the de~ree of puriflcation obtained.
Extensive washln~ !n the presence of stabilizers plus hl~her salt and d'lfferin~i deter~en~s may be utilized to remove nonspecifically absorbed proteins. Elution may then beadvantaçjeousiy carried out either usin~ specific peptide elution if one has antibodies to Ci TR peptides. (Courtnei~e et al., Cold Sprin~ Harbor Conf on Cell Prolif. and Cancer 2123 (1984)), or altnerativeiy by chaotropic agents such as potassium thiocyanate or by bwerin~ the pH followed by immediate pH neutralkation of the eluted fractlons.
IG4-9.2 Although it is likely that immunoafflnity chromatography would provide a si~nificant purification and provide protein of sufficient purity for research studies and dru~ screening, such an approach alone may not provlde adequate protein pur'lty to qualify the CFTR protein as a clinical ~rade therapeutic agent. Thus, to pur'lfy the protein further. or in the case that immunoaffWty chromatography was unsucce$ful, one could evaluate additional chromato~araphlc approaches to select an optimal chromato~raphy procedure to obtain the desired purity. For example, Iigand affinity (Landry et al., Science 244 1469 (1989); Smlgel, J. Blol. Chem.261 1976 (1986)), lectin (Curtis and Catterall, Biochemistry 23 2113 (1984)), anlon exchan~e (Hartshorne and Catterall, Proc.Natl. Acad. Scl. 78 4620 (1981)), hydroxylapatite (Hartshorne and Catterall, J. Biol. Chem. 259 1667 (1984)), and ~el flltration (Borsotto et al., J. Biol.
Chem. 260 14255 (1985)) chromato~raphy procedures have been used in puriflcationschemes for this class of membrane bound proteins. Since the cFrR protein contains a nucleotide bindin~ domain, it woill likeiy bind to resins such as Cibicron blue and may be speciflcally eluted with nucleotides (Lowe and Pearson, Methods in Enzymology104 97 (1984)). The accessibility of the nucleotide bindin~ domain in the solubilked form would have to be determlned empirically, The predicted protein sequence forthe CFTR contains a carbohydrate attachment sHe at amino acld 894, Since it has now been shown that the CFTR protein is a ~Iycoprotein, the use of lectin 2 0 chromatos3raphy is a likely route to purify Ci-TR, Example 11 - Preparation of CFTR Protein Specific Antibodies Monoclonal antibodies MAb 13,1 and MAb 13.2, specific for predetermined 2 5 re~ions or epitopes of the CFTR protein, were prepared using the following cloning and cell fuslon technique. A mouse was Immunized with the polypept~de produced from Exon 13 of the CFTR proteln fused to i~aloctosldase, the fuslon proteln being obtained as described in Mole and Lane, DNA ClonlnS~ Volume lll: A Practlcal Approach (1987), to induce an immune response. Tine immunizatlon procedure requlred InJectin~ a mouse with 1û micro~rams of Immuno~en In 10 mlcroliters of PBS
emuisified In 30 microliters of Freunds complete adJuvant (Glbco #660 5721AS). This procedure was repeated four times at intervais of from 1 to 28 days over a 57 day period. The mouse was then In~ected with 50 mlcro~rams of Immuno~en In 5û
microl'rters of PBS four times over a three day period. Vasodilation was induced by warming the mouse for 10 minutes with a desk lamp. The mouse was sacrificed by C2 intoxication and a splenectomywas performed.
After immunr7ation was carried out, the ~-iymphocytes of the immunized mice were extracted from the spleen and fused w'lth myeloma celis using the well known processes of Koehler and Milstein (Nature,256 (1975),49~497) and Harlow and Lane, Antibodies, A Laboratory Manual, Cold Spring Harbor Laboratory, New York (1988),respectively. The resuiting hybrid ceiis were cloned In the conventional manner, e.g.
o using limiting dilutlon, and the resuHinçl ciones, which prociuce the desired monoclonai antibodies, cuitured. Two most preferred antibodies produced by this process were MAb 13.1 and MAb 13.2, specific for i-xon 13.
The monoclonal antibodies, MAb 13.1 and MAb 13.2, may be used in their complete forrn or as fragments thereof (e.g. Fab or F(ab')2 fragments) providing they exhibrt the desired immunological reactiv'lty w'lth CFTR or the desired Ci-TR domain.
The term ~monocional antibody- as used herein therefore aiso includes such fragments. The monoclonal antibody is ideaily used in an immobilized form, and is most preferably immobil'eed on a resin substrate, for purification of the Ci TR protein from other contaminants. The antibodies can also be advantageously used as part of a kit to assay for the presence of the CFTR protein in biologicai samples such as fluids or on the surface of ceiis.
Hybridomas producing monoclonal antibodies MAb 13.1 and MAb 13.2 prepared according to this procedure have been deposrted w'rth the American TypeCuHure Collection (ATCC) under the terms of the Budapest Treaty, and assigned accession numbers: ATCC 10565 and ATCC 10566.
IG4-9.2 ~ r~ !~ r7 ;~
Example 12 - CFTR Production Results from Cells Transrormed with Various CFTR aenes includin~ Munants A. CFTR from T84 cells. Previous examples show that CFTR can be detected in T84 celis by adding (~3~P)ATP and protein kinase A to immunoprecip'itanes formed using antibodies raised aS~ainst CFTR (see aiso Gregory et aL,1990). Band B, and largeamounts of band C were detected by thls method (see Fi~ure 9). Partial proteolysis fingerprinting showed that the T84 cell derived material and that produced in a celi-free system directed by Ci-TR RNA were Ind'lstingulshable.
Figure 9 demonstrates that band C is CFTR modified by addirtion of N-linked carbohydrcne. Upon treanment with N-GLYCANASE(R) enzyme, band C, immunoprecipitated from T84 cells and phosphorylated in vitro, is converted to band A. Treatment w'ith ~GLYCANASE(R) enzyme, endo~lycosidase H or endogiycosidase F enzymes had no effect (Fig. 9). Because a band of Intermediate molecular weight was aiso detected upon treanment with N-GLYCANASE(R) enzyme, these resuHs can be interpreted to mean that CFTR bears two complex carbohydrate side chains possibly of the tri- or tetra-antennary type. N-GLYCANASE(R) enzyme treatment ofband B aiso yielded band A (Fi1. 9) (see also Gregory et aL,1990). The shift inapparent molecuiar weight on polyacrylamide gels in goinJ from band A to band Cseems iarge (20K) bun whether this represents addHion of unusually large s~de chains, or mereiy resuHs from anomalous mlgranlon In SDS-polyacrylamide geis 'Is unknown. H is postuiated that siycosylation of band C is probably also responsibie for Hs migration cs a diffuse band as opposed to the sharp appearance of bands A and B.
B. ~F508 does not Produce Mature CFTR. Recomblnant CFTR has been expressed util'rzins~ a vacclnia virus-lnfected HeLa cell system (see aiso Grejory et al.,1990; Rich et aL,1990). Because of the short infectlon cycle of vacclnla vlrus, longer termexpresslon was studied In transfected COS-7 celis (see Example 7). WHh reference to Figure 10A, COS-7 ceils were elther mock transfected (lane 2), transfected wHh wild type CFTR (pMT-cFrR - lane 3) or the mutants pMT-CFTR-~F508 (lane 4) and pMT-CFTi~
Tth 1111 (iane 5). Lysates were prepared 48 h post-transfection, phosphorylated in vHro IG4-9.2 wHh proteln kinase A and (~32P)ATP and anaiyzed on a SDS-polyacrylamide ~iel. Lane 1 contains iysate from T84 celis. The positions of bands B and C are indicated on the ri~ht margin. Autoradio~iraphy was for 2 h. WHh rsference to Figure 10B, the 32p In vitro labeled bands C from T84 cells (lanes 1-3) and from COS-7 celis transfected wHh s pMT-CFTR (ianes 4-6) and band B from cells transfected with piviT-CFTR (lanes 7-9) we!e excised from the gel and dlgested wHh increasing amounts of S. aureus V8 protease.
Proteins in lanes 2,5 and 8 were dlgested with 0.017 iug/;ii of S. aureus V8 protease ar.d those in lanes 3,4 and 7 with 0.17 ;lg/ili of enzyme. Lanes 1,6 and 9 were untreated samples. Exposure time was two days.
Thus, Figure 10A shows Ci TR produced in celis transfected with an expre$ion plasmld (piviT-CFrR) containing a full length CFTR codin~i sequence expressed from a mouse metallothionein promoter. Using the 32p in vitro labeling technique and affinity purified polyclonal antibody to exon 13 fusion protein (see also Examples 10,11,17 and also Gregory et al.,1990), band C was readily detected in transfected cells, as well as smaller amounts of band B (lane 3). COS-7 cell band C migrated more slowly than the CFTR from T84 cells (lane 1) but figure 10B shows partial proteolysis fingerprints that confirm that the proteins are indeed related. Presumably, the glycosylationpattern of human colon and simian kidney celis is sufficiently different to alter the 2 0 mobilHy of band C.
Figure 10A also shows that COS-7 cells transfected wHh vectors containing a ~F508 cDNA produced band B but, unexpectedly, they did not contain band C (lane 4). Similariy, a mutant CFTR truncated by Insertion of a frame sh'fft mutation at the Tthl 111 site (whlch resuited in the synthesis of a 1357 amino acid protein) encoded a truncated version of band B of predicted molecular wei~ht but also lacked the band C equlvalent (lane 5).
To conflrm this data, metabolically labeled COS-7 celis were used. After the celis were labeled wHh (35S)methionine for 16 hours, they were Iysed and ImmunoprecipHated wHh monoclonal antibody mAb 13- 1 (raised against exon 13 fuslon proteln) (see Example 11). Flgure 1 lA shows that band B was labeled in COS-7 celis transfected with wlld type (lane 2) and ~F508 cDNA (lane 3) but surprislngly, that labeled band C was totally absent in the mutant cDNA transfected cells.
IG4-~.2 ~ jf~ r~ ~
Fi~ure 1 1 B shows the resuH of a pulse-chase experiment in which COS-7 cells, transfected with wiic type and ~Fso8 cDNA vectors pursuant to i~xample 7, were labeled for 15 mins and chased over a 24 hour period. Wild type band B chased into s band C such that by 4 hours after labelinç~, very little band B remains (lane 4). Mature CFrR was observed at 1, 4 and 8 h post labelln~a but by 24 hours, little remainin~j labeled material was detected. By contrast, aHhou~h ~F508 band B was metabolizedwith approximateiy the same half-life as wild type, no band C appeared.
Not all labeled band B In puise labeled wild type cDNA transfected cells appeared to be processed to the fully ~iycosylated band C. One interpretation of this flndin~ is that recombinant celis contained such lar~e amounts of cFrR that the machlnery responsii~e for further post translational processin~ was saturated. Under these circumstances, excess material may be de~raded. An aHernative explanation is that durin3 the chase period, so much unlabeled CFrR accumulated that insufflcient antibody was present to capture all the labeled protein. Studies wHh vacclnia virus-infected HeLa celis synthesizin~ CFrR showed that very rlttle band C material was detected in a 1 h labelin~ period. This lai~eling pattern is consistent with the kinetics shown here.
C. Immunofluorescence Studies. rhe absence of mature cFri? In ~F508 cDNA
transfected COS-7 cells implies that the deletion caused a structural aiteration that ~ .
somehow preventec maturation of the carbohydrate in the Gol~i. rhis could resuH
because transpon from the endoplasmlc retlculum to the Gol~l was inhibHed or 2 5 because modiflcation was inhibHed even thou~h transport was normal. It was hypothesized that if protein transporl were Inhibited ~t miç~ht be possible to detect a difference In location of mutant and wild type recombinant CFTR by Immunofluorescence.
Fi~ure 12 shows Immunofluorescence photomicro~raphs of COS-7 cells transfected with wi~ type and ~F508 CFrR cDi~iAs usln~ monoclonal antibody mAb 11. That the fluorescence detected was CFrR Is indicated by the previous characterization of t )e monoclonal antibody, by the absence of si~nal in non-transfected celis (back~round celis In Fl~. 1 2c and 1 2d) and because the reactlon was InhibHed by exon 13 fusion protein (Fl~. 12b) but not irrelevant fuslon protein.
IG4-~,2 Fl~ures 12c and 12d show that the su~cellular distributlon of wild type and ~F5~8 CFTR
was different . The ~Fso8 si~inal appeared localked to the perinuclear re~ion whereas the wild type CFTR si~nal was more diffuse. The pattern observed with wild type su~ests a wide-spread distribution possibly includinS~ the plasma membrane.
i3ecause the distribution of CFTi? In recombinant celis overexpressing the protein may not be typical, subcellular localkation of wild type and ~F508 was not reflned.
Subcellular distribution of ~F508 Ci-TR was different from wild type.
D. Other Mutations Prevent Maturation of CFTR. To study the maturation of CFTR in more detail, additional site specific mutations wHhin the cDNA codin~ sequence were constructed. A naturally occurrin~ deletion mutation at residue 507 was created by removin~ the codon for isoleucine (Kerem et al.,1990). To examine the role of nucleotide bindin~ wHhin the domain includin~ ~F508, the hi~hiy conserved Iysine at residue 46~i (Riordan et al.,1989) was chan~ed to methionine. The equivalent mutation was also made wHhin the second nucleotide bindinç~ domain (K1250M) and both aspara~lne residues (at 894 and 900) were chan~ed to ~iutamine to which carbohydrate is predlcted to be attached (N894,900Q)(Riordan et ai.,1989).
Vectors containin~ each of these mutations were constructed and separateiy transfected into C0~7 cells. WHh reference to Fl~ure 13.
expression vectors containin~ wild type CFTR (pMT-CFTR - lane 2) and those containing the mutants pMT-CFTR-K464M (lane 3), pMT-CFTR-K1250M (lane 4), pMT-Ci-Ti?-~i507 (lane 5), pMT-CFTR-NB94,900Q (lane ~, marked as pMT-CFTR-deglycos.) and pMT-CFTR-R334W (lane 7) were transfected Into C0~7 celis. Lane 1 is C0~7 cells whlch had been mock transfected. Lysates were prepared 48 h post-transfection and the immunopreclpitates formed usin~ pAb i x13 were labeled in vHro usin~ protein kinase A and (~32P)ATP. The posHions of bands A, B and C are indicated on the right margin.
Autoradio~raphy was for 2 h.
Fi~ure 13 shows that usin~ the in vitro kinase a$ay, ~1507 cDNA transfected celis, like their ~F508 counterparts, lacked band C (lane 5). N894,900Q produced neHherband B or C, but Instead ylelded a band of sli~htiy increased mobiiHy which was interpreted to be the CFTR primary transiation product, band A, of apparent 1G4-9.2 . .
Q 3 ~
molecular wei~ht 130kd (lane 6). This confirmed th~t it was the addition of N-linked carbohydrate to CFTR that caused the mobllity shifts resultin~ In banâs B and C.Ind~idual mutations in each of the two SneS was requlred to establish unequivocally that iooth Asn894 and AsnsO0 are glycosyiated and based on the N-GLYCANASE(R) enzyme resuits, this seems likeiy.
K464M cDNA transfected celis, like their ~507 and ~F508 nucleotide binding domain 1 mutant counterparts, contained no band C (lane 3). Surprisingly, however, the equivalent mutation In the conserved ysine of the second nucleotlde binding domain did not prevent maturation (lane 4). Another rare but naturally occurrinamutatlon associated with CF occurs at resldue Ar~334 withln transmembrane domain6 (X. Estivill, personal communication). This mutation, R334W, did not prevent maturatlon of recombinant Ci TR band C, (Lane 7).
Table 2 summarizes data obtained with all the mutants Including two other naturaliy occuring CF associated mutations S549i and G551D. These were from a second cluster of mutations within the first nucleotide binding domain, in this case within exon 11 (Cutting et aL,1990a, Kerem et aL,1990). Also included is F508R, in which the residue at 508 was changed rather than deleted. Surprisinaly, the resuits 2 o using these mutants showed S5491 Ci-TR does not mature but G551 D does. The mutatlon of phenylalanine 508 to arginine also resuited In CFTR that did not mature.
Example 13 Intracellular Characterizatlon of Ci TR
A. Endoplasmlc reticulum interactions. i3ased on the discoveries of thls invention, nascent CFTR interacts first with the endoplasmic reticulum and Is then glycosylated at at least one of Asn resldues 894 and 900. The native molecule is then transported to the Golgl where carbohydrate processlng to complex-type giycosylation occurs.
Finaliy, at least some of the mature glycosylated molecule is thereafter transported to the plasma membrane.
IG4-9.2 `.3 1 ~ J ~ ) It Is now reasonably well established that the endoplasmic reticulum possesses amechanism that prevents transport of mutant, misfolded or Incorrectiy complexed versions of proteins otherwise destined for further proce$ing (Lodish,1988; Rose and Doms,1988: Pelham,1989; Hurtley and Helenius,1989 Kiausner and S'rtia,1990). If this qual'riy controi mechanism operates on CFTR, H would prevent transport to the Golgi and consequentiy, further modification of several of ihe mutants reported here. As a resuit, the unmodified mutant versions of the protein eHher would not exH the endoplasmlc reticulum and would subsequentiy be de~raded therein, or alternatively, they would be transported to the iyosomes for degradation.
It is not clear how the quality control mechanism recog;nrzes the d~ference between wlld-type and those mutant versions of Ci TR whlch were not further processed. One obvious mechanism would be that an aiteratlon in structure of themolecule Is detected. Indeed, gro$ changes in structure of the flrst nucleotide bindin~ domain (and perhaps in consequence of the whole molecule) might be expected following deletion of phenylalanine 508 (Hyde et al.,199C; Manavalan and Dearborn, personal communication). However, H is not clear how this change in structure would be detected by a mechanism located, for example, in the lumen of2 0 the endoplasmic reticulum, since the domain bearin~a the mutation, (if the present model for Ci~ is correct), would lie on the cytosollc side of the membrane. Perhaps the structural change is transmr~ted across the membrane or perhaps the sensing mechanism does not reco~nke Ci~r~ directly, but rather detects a protein wHh which it Is complexed. In this case, all mutations wHhin CFTR that prevent complex formation aiso prevent Intracellulartransport. Yet another mechanism would be that nascentCFTR has basal activHy in the endoplasmic retlculum and that mutatlons that disrupt this actlvHy are sensed by the qualHy control mechan~,sm. Perhaps some activ'Hy of CFTR is necessary for its maturation and by this means, enzymatically inactive proteins are mari~ed for degradation. Irrespective of the mechanlsm of d~iscrimination, the time course of synthesis of both wild type and mutant CFTR is notable in two respects. Firstly.
the half life of band B is similar for both wild type and mutant versions and secondly, most of the wild type band B appears to be de~jraded. One Interpretation of these resuits is that synthesis of CFTR invoives two steps, retention in the endoplasmlc IG4-9.2 - . - ~ .
' i) `J
retlculum during which time foldin~a of the protein occurs followed by eHher export to the Golgi or degradation. Since we detect no difference In the resldence time in the endoplasmic reticulum, H would appear that the defect In the case of the non-maturin~a mutants rleS in tile second step, that which resuHs in de~jradation.
Furthermore, even wild type seems surprisinS~ly susceptible to de~radatlon since most of band B fails to mature to band C. Whether thls results from overexpre$ion of CFTR o-is a property of the protein In non-recombinant celis remains to be determined.
Still alternativeiy, the CFTR protein Hseif may be responsibie for Hs own exportation out of the endoplasmic reticulum. Under this Interpretation, mutant CFTR, or otherwise improperiy folded or ~iycosyiated CFTR would not appropriately interact wHh the endoplasmic ret~lculum membrane resuHing In a seif-re~aulatinçl quality control mechanism havin~ no need of further structures or accessory sui~stances.
A different interpretation of the resuits would provide that the nascent, incompleteiy ~iycosyiated CFTR was transported normaliy to the Golgi but that the structural alterations caused by the various mutations prevented further glycosylation and this lead to lack of activHy and eventual de~radation. This interpretation is less favored because the prevlous explanations are more conslstent with the present 2 o understandin3 of the intracellular transport of other proteins and their mutant variants (Lociish,1988: Pelham,1989; Klausner and SHia,199û).
B, Structure:Functlon of CFrR. CFTi~ is a lar~e, complex molecule. Nucleotide bindin~ domain 1 contains two clusters of naturally occurinçl mutations, one around residue 508 (Riordan et al.,1989, Kerem et al.,1990), the other around 550 (Cuiting et _.,1990a: Kerem et al.,1990), All the mutatlons around 508 disclosed herein (~F508, Q1507, F~08R) falled to ~enerate mature CFTR, whereas mutatlons at the second sHe, S5491 dld not produce mature CFTR but G551 D dld. Mutation of the Walker motif iysine In nucleotlde blndln~ domaln 1 aiso prevented maturation of CFTR. The surprisingdifference between mutatlons at neighboring resldues 549 and 551 is a surprisin~ resuH.
It appears that most of these mutatlons Inactivate some functlon of the protein, such IG4-9.2 r) ~1 7 o as Hs ability to bind nucleotide and or maturanion of CFTR Is prevented by lack of functional activity. More likely, all non-maturin~ mutants resuit In structural chan~es in the domain and these prevent manuration~
s Another unexpected result of the experiments disclosed herein is the difference between the mod~icanion of the conserved iysine mutants in nucleotide bindin~
domains 1 and 2. K464M did not produce mature CFTR whereas K1250M did.
Althou~h the two domains are cleariy related and both mutations iie in punafive nucleotide bindin~ pockets (Riordan et al., 1989), they appear not to be functionally equivalent.
Munant R334W (x~ Estivill, personal communlcation) emphasized the importance of the transmembrane domains in the activny of CFTR. The Instant disclosure clearly shows that a chan~e in sequence within transmembrane domain 6 does not prevent movement to the Golgi at least as measured by the presence of complex-type N-linked oli~osaccharides. Accordin~ly, the polar amino acid in the otherwise hydrophobic environment plays an important roie in pumpin~ materiai across the membrane.
Exampie 14 - Cvstic Fibrosis Disease implications - Diaanosis and TheraPy A. Molecular basis of the disease. Many ~enetic diseases are caused by the absence or truncation of the appropriate protein, for example as a result of deletions wrthin the conespondin~ ~ene. Muscular dystrophy would be an example in this cate~ory (Harper, 1989). Other ~enetic diseases are caused by munations that resuit in loss of functlon of the ~ene product. Sickle cell disease is a classic example of this type (weanherall et d., 1989). One aspect of the instant invention provides than the molecular basis of most cystic fibrosis is the inability of the CFTR ~ene product to mature, That is to say, the fallure of CFTR to move throu~h the nonmal panhway of 3 o intracellular traffickin~ and modificcnlon means that the mature protein is absent from nS flnal cellular destlnation In CF celis. Examples of S~enetic leslons that result in failure of the LDL receptor to mature have been described for certain types of famiiial hypercholesterolemia. In some of these cases, the mutant LDL receptor 'IS retained In the endoplasmlc reticulum (Lehrman et al., 1986).
IG4-9.2 That little or no mature CFTR has ioeen detected In the celis containin~ CF
assoclated mutations ooserved in a majority of CF patients does not necessarily mean that this forms the molecular bas'ls of all CF. A priori, it seems very likely that some mutations will Inactivate the function of CFTR but will not prevent transport and glycosylatlon. Indeed, R334W and Gss1D have been detected In CF chromosomes and presumabiy encoded inactive cFrR (X. Estivill, personal communication; Kerem et al.,1990). Even so, both encoded CFTR that matures to fomm band C.
B. Dia~nos'ls. The mutatlons descrii~ed herein represent over 70% of known CF
chromosomes (Kerem et al.,1989,1990; Riordan et al.,1989; Cuttln~ et al.,1990a).Accordin~ly, the suri~risin~ resuits of the instant Inventlon can be used for purposes of dia~nosinç1 CF. Furt~er, it is anticipated that mutations In other CF chromosomes will aiso fail to produce icand C, thus makin~ the detectlon of CFTR protein In the membrane diaS~nostic of an even S~reater percenta~e of CF. Another aspect of thepresent invention is the dia~nosis of CF by monitorinçj the presence or absence of mature CFTR. Accordingly, the sensitlve detection of band C in primary cells provides a surprisin~iy useful diagnostic test for detectin~ the ~reat maJority of CF patients.
2 0 C. Pancreatlc sufficiencv and InsufficiencY. To date some mutations that cause premature termlnation of CFTR synthesis appear associated with mild forms of CF,whereas ~F508 is often assoclated with severe, pancreatic Insufficlent forms of the diséase (Cuttln~ et al.,1990b). That ~F508 should be more severe than a maJor truncation appears counter intuitive. The experimental data disclosed herein support the conclusion that rnaJortruncotlons make no stable CFTR. By contrast, homozy~ous ~F508 cells not only make no moture CFTR but worse, they produce mutant protein trapped in the endoplasmlc retlculum. Trapped ~F508 CFTR may retaln sufficient activity to cause Intracellular pumping of molecules normaliy transported only at the cell surface. Thus, cFrR activity at the Incorrect cellular locatlon would resuH in effects more serlous than those resuHlng from complete absence of the protein. Accordln~ly.
suitable therapeutlc activity would Ideally deactivate such Inappropriate CRTR activny most preferably, In advance of, or In conJunctlon with CFTR protein or CFTR ~enetherapy.
IG4-9.2 ~ 93 C~ 7 L~
D. Recessive nature of CF. ~he absence of mature CFrR encoded by ~F508 and other simllar mutants aiso provides an explanation for the findln~ that ceiis heterozygous for various mutations are apparently wild type in cell surface channel activities associated wHh CFTR. Previously, n was perhaps surprising that the defective s molecule did not interfere wlth the activity of the wild type. From the Instant invention, it was surprisin~ly discovered that celis heterozy~ous for ~F508 completely lack mutant Ci^TR at the cell surface and in consequence, the wild type protein Is able to function uninterruptedly.
0 E. Therapy. The Instant discovery that the majority of cases of CF are caused by the absence of mature CFTR and possibly, in the case of pancreatic insufflclency, by the addHional deleterious effects of Incorrectiy located, partlally active CFTR, confirms the basis of other approaches to CF therapy. For example, drugs active in aitering the sui~cellular distribution of proteins could advanta~eously ioe used to redistribute to the plasma membrane fully ç~lycosylated mutant forms which retain at least some functlonal activity. Similariy, agents effective in simulating sufficient Ci-TR activity to resuit in export of otherwise mutant Ci rR to the Golgi for addHional ~iycosylation could resuit in Improved CFTR functlon in homozygous CF individuais. Aiternativeiy, therapeutic treatment via a suitable, therapeutically effective blocking agent could 2 0 be used to deactivate inappropriateiy located, active, mutant CFrR protein.
Aiternately, one may promote the transport of such protein to an appropriate location and useful In this regard are rea~ents active in promoting intraceiiular transport inhibitlon. Yet another aspect of the present invention reçJardin~a the therapeutic treatment of mislocated CFTR comprises the use of anti-sense nucieic acid to rid cells of mutant transcript to provide the absence of CFrR which is preferable to incorrectly located proteln.
IG4-9.2 Most preferably, treanment of Indlvlduais wnh CF will comprise the administration of a therapeunically effective amount of replacement CFTR protein. Ideally, the CFTR
will be administered via aerosol Inhalanion so that it is applied directly to the airway s celis. The CFTR protein could be formulated In a lipd containing vehicle such as liposomes or in virosomes. The flnal formulation will advantageously comprise a carrier as a vehicle for physically transporting the most preferred embodiment will alsocomprise a dissolvl'n~ agent for dissolvin~ the mucous or otherwise assisting the movement of the cFrr through the mucous layer to the airway cellular membrane.
o Ideal rea~ents In this re~ard would tar~et the CFTR and/or the delivery vehicle to airway celis and further, promote fusion therewnh. Reagents active in this manner include viral proteins such as the HA protein (for tarç~etlnçj) and F protein (for fusion) of parainfluenza viruses.
Example 15- Formulation of CFTR Protein into Artiflcial Uposomes Solubilked preparanions of CFTR, whether or not purifled, can be reconstnuned into artiflcial liposomes (Klausner et aL, In Molecular and Chemlcal Characterizanlon of Membrane Receptors Alan R i iss N.Y. ~1984) p209). Detergent solubil~eed preparatlons 2 o of CFTR can be added to phoshollpld suspens~ons and the detergent removed, and vesiculanlon induced either by dialysis (Kagawa Y, Kandrach et aL, J. Biol. Chem. 248 676 (1973)), chromano~raphy over Sephadex G-50 (Brandt and Ross, J. Blol. Chem. 261 1656 (1986 or by passing the preparatlons over i:xtracti-Gel D (Feder et al., EMBO J. 5 1509 (1986); Cerione et aL, J. Biol. Chem. 261 3901 (1986)) or by other methods known to one skilled in the art. For example, forthe bovlne adenylate cyclase, Smigel (Sml~el, J. Bioi. Chem. 261 1976 (1986)) found that the cyclase could be reconstnuted Into llposomes by passlng a solutlon contalnln~ CHAPS bu~fer solubllized cyclase,1.5 mM phosphatidylethanolamlne and 1.0 mM phosphanldylserine over a Sephadex G-50 column. Naturaliy, obvious experiments also can be carried out to detenmlne the optlmal llpld composition of the artiflclal llposomes needed to achieve fusion or Implantanlon of CFTR Into CF celis. In ~eneral, membrane proteins orient themselves conectly In llposomes (Klausner et aL). The conect orientanlon can be determinedusln~ antlbodles, and if necessary, the separanlon of conectly-oriented from Incorrectiy-oriented llposomes can be achleved usln~ Immunoafflnity chromano~raphy (Anhoit et al., J. Blol. Chem. 256 4377 (1981)).
IG4-9.2 ~ 'ifi ~
Example 16 - Gene therapY
A ~enetic therapy approach to treatment of cystic flbrosis would make use of the full len~th cDNA encodin~ the CFTR to au~ment the defective ~ene and ~ene product. Thls approach could entail either Introduction of the CFTR cDNA capable of expression of CFTR into human cells In v'rtro followed by transfer of the cells into the patient or aiternatively, one may directly introduce the CFTR cDNA containin~ vectors into the cystic fibrosis patient. cDNAs recentiy have been introduced successfully into 10 humans by Rosenber~, Anderson and collea~ues (Aebersold et al., J. Cell Biochem.
Supplement 14B,78 (1990)).
Current ~ene therapy approaches are based on the use of modifled retroviral vectors for the introduction of protein codin~ sequences into cells and animals. For example, using the full len~th CFTR cDNA of the present invention, similar techniques can be applied to introduce CFTR coding sequences into cystic flbrosis patients.
For example, Um et al. (F~roc. Natl. Acad. Sci. 86 8892 (1989): Mol. Cell. Biol. 7, 359 (1987 descfibed techniques and vectors for a ~ene therapy approach to 2 0 expression in vlvo of the human adenosine deamnise ~aene in hematopoetic stem celis. This system could be easiiy modified to provide for a S~ene therapy approach to in vivo expression of the CFTR protein. The work of Rich et al. (1990) and Drumon et al.
(Cell 62,1227 (1990)) conflrms the feasibility of thls approach.
2s Additional limitations and criteria re~ardin~ the control of CFTR expression followin~ ~ene therapy will also become apparent upon study of the resuits of protein productlon from the variou5 mutants and the manner In whlch nascent CFTR interacts with the endoplasmic reticulum, transported to Gol~i for further carbohydrate processin~ and subsequent transport to the plasma membrane. Examples 12 and 13 30 are particulariy helpful In this re,~jard.
It is now clear from the present invention that ~ene replacement therapy for CF
will need to controi strictiy the level of expression of CFTR because overexpression will saturate the transport system involved in maturation. Additionally, CFTR mislocated by over-expression could be as deleterious as protein mislocated by mutation.
IG4-9.2 . .
.. ~ ~ ' - . :
. : , : ~ . -" . ' ' '. ~
Accordingly, the protein replacement therapy is preferred since such an approachadntageously avolds this hazard.
Example 17 - Dru~ Screenin~ for Pharrr~acoloaical apents A pharmacolo~ical approach to develop CF therapies would use cells expressing CFTR from the DNAs of the present inventlon to screen for and select agents, eHher natural products, recornbinant products or synthesked organlc moiecules, that could i~e used therapeutically to compensate for or by-pass the defective CFTR. For example, ionophores capable of aitering membrane conductance or lon channel agonists or antagonlst could be potentially useful compounds. Aiternativeiy, agents for mobilizing mutant forms of CF~ to the golgi for glycosylation to partialiy active cFrR for CF patlents could be isola1ed.
To test for potential pharmaceutical agents, the cell systems of the present inventlon, eHher expresslng wild-type or mutated forms of CFr ? protein from the full len~th cDNA or isolated DNA sequence encodinS7 CFT ?, would i~e Incubated in thepresence of varyin~a concentrations of the agent beln~ tested and restoratlon of the wlld-type phenotype or blndlng of the agent to the cell or CFT ? assayed. An example of a suHable assay fortesting the restoration of appropriate ion flux, has been described In detall by Mandel, J. Biol. Chem. 261,704 (1986) and Clancy, Am. J.
Physiol., 258 Lun~ Cell. Physlol. 2 pL25 (1990). Aiternat~ely the detecting step could comprise contactin~ the celis wHh a iai~elled antii~ody specific for the cystic flbrosis transmembrane conductance reguiator and detecting whether the antibody became bound wherein binding is correlated wHh the presence of an effective agent For screenln~ molecules as potential CF therapeut~c drug candidates, one could assess the effect of exogenous materials on the function and phenotype of cel~
expressing enher wild-type or defective CFTR. One could examine the cr transportproperties as descrii~ed by Mandel et al. U Blol Chem 261,704 (1986)) or one could use the measurement of 1251- efflux (Clancy et ai., Am. J. Physlol.258 Lung Cell.
Physiol. 2 pL25 (1990)).
1G4-9.2 ,; ' .
, ~ ~ ~i 7 ' 7 3 Measurement of 125l- effiux from Intact celis provides a relatively easy and fast assay of cr channel activ'riy. I- is an excellent tracer for cr: 'rt is not secreted across the epithelium (Widdicombe and Welsh, Am. J. Physiol. 239, C112 (1980)) but both the secretago~ue-induced apical membrane Cl- conductance and the outwardly rectifying apical cr channel are more permeable to 1- than to cr (Li and Welsh, Clin.
Res. 37,919a (1989)). Dr. Welsh and collea~aues have shown that 1251- effiux: a) is stimulated by an increase In cAMP, by an increase in Ca2+, and by cAMP and Ca2+
elevatlng a~onists, b) is inhibited by carboxylic acid analo~s, c) is not affected by loop diuretics, and d) Is voita~dependent. These data indlcate that the 1251- efflux assay measures cr channel activriy.
The resuHs of various mutant CFTR expressin~ celis at 5~75% confluency at amblent CO2 and room temperature (20-23C) is described in prior examples. Cell attached cr channeis have a slmilar function at room temperature and at 37C. For testln~ the effect of varying concentrations of substances on the CF phenotype, one could Include the substances in the preincubation media and then subsequently conduct effiux measurement assay. Followin~ preincubation one would remove the media, and cells would be washed for 10 seconds in efflux buffer containing (in mM):
135 NaCI,1.2 CaCI2~ 1-2 M~CI2~ 2 4 K2HPO4~ 0-6 KH2PO4,10 ~llucose, and 10 HEPES (pH
2 o 7.4 with NaOH). Celis would then be loaded w~rth tracer by incubation in buffer contalnir)g 15 iilCi/ml 125r for 2-4 hours. Celis then would be washed for 30 sec to remove most non-specifically ioound tracer thereby producin~ a stable baseline rate of efflux. 1251- effiux rates could be measured durin~ a basellne perlod (5 minutes) and then during stimulatlon wrth either cAMP (100 jilM CPT-cAMP,10 jlM forskolin, and 1 mM theophylline) or Ca2+ (1 iuM A23187 or 1 jlM ionomycin). Measurement of efflux In response to a Ca2+ lonophore would provide an important control because an increase in Ca2+ activates cr channels in CF celis. Efflux buffer from all time periods plus non-effiuxed (iysis) counts would be quant~itaned In a çlamma radlation counter.
To Increase the unility of this method, the procedure could be adapted to cells ~rown in 9~ well dishes.
Aithough Impractlcal for wide spread druç~ scr~enln~, In order to further characterr~e promisina candldane molecules, patch clamp studles could be performed on wlld-type or mutant CFTi~ expressin~ cells. Methods for cell-attached IG4-9.2 - 2~7~;7 ~
and excised, irlside-out pcnch clamp studies have been described (Li et al., Nanure 331,358(1988):Weisn~sclence232~l648(l986))~ Crchannelswouldbeident'lfiedby their size, selectivity and characteristic outward rectlficatlon. With cell attached patches the effect of substances under study could be examined by their addition to the bcnh~ wnh excised patches the effect of adding substances to the cytosolic surface or external surface of the patch could be determlned. Uslng these assays, promisinçi lead cornpounds for the treatment of CF could be Identified.
It would be advantageous to develop addHlonal rapld assays for monitoring the CFTR proteln. Aithough the exact function of the CFTR protein is not known, the presence of nucleotide binding domains of other protelns suggests that the CFTR may react wHh radlolabeled nucleotlde analo5~ues or could hydroiyze nucleotlde triphosphanes~ Fa example, attempts to photoaffinity label CFTR Wnh 8-azido-c~-(32P)ATP could fobw the baslc protocol of Hobson et al. (Hobson et al., Proc. Natl.
Acad. Scl. 81 7333, (1984 as successfully modified for labelinçi of the multi-drug resistance, P giycoproteln (Cornwall et al., FASEB J 1, 51 (1987)). Membrane vesicles from cells or solub~ed micelles could be incubaned In HEPES buffered mannnol wnhMnCi2, MgC12 and photoaffinity label. Samples would be irradiated at 366 nm and then eHher electrophoresed directly on SDS geis to determine the extent of labeling or 2 o immunopreclpHated to quantHate label Incorporated Into CFTR.
AddHionaliy, one could advantageousiy attempt to measure ATP hydrolysis by modificatlon of the procedure used by Hamada and Tsuro for measuring the ATPase activity of P-giycoprotein (Hamada and Tsuro, J Biol Chem 2~3 1454, (1988)). Ci-TR
could be solubllized as disclosed and Immunopreclpltated by reaction wHh antibody and then proteln A-Sepharose followed by Incubatlon In the presence of (c~-32P)ATP.
The reactlon wouid be stopped by the addition of EDTA and excess nonradioactive ATP and ADP. The reactlon products would be separated by chromatography on poiyethylenelm~n~cellulo5e thln layer plates, the ADP-containing spots detected by UV ll~ht and quantHated (Cerenkov), Qualitatlve hydrolysls could be determined by autoradlography of the TLC plate. In drug screeninçi, the effect of varying concentrations d added substances on these assays could be determined and molecules wHh potential as CF therapeutlcs Identified.
~ ~ ~ r~ 3 Cuttin~, G.R., Kasch, L.M., Rosenstein, B.J., Tsui, L-C., Kazazian, H.H.Jr. and Antonarak'ls, S.E. (1990b). Two cystic fibrosis patients with mild pulmonary disease and nonsense mutations in each CFTR ~ene. Arn. J. Hum. Genet. 47,213.
Dean, M., White, M.B., Amos, J., Gerrard, B., Stewar-, C., OSKhaw, K.-T., and Leppart, M. (1990) MuHiple mutations in hiç~hly conserved residues are found ~n mildly affected cystic flbrosis patients. Cell 61,863-870.
Drumm, M.L., Pope, H.A., Cliff, W.H., Rommens, J.M., Marvin, S.A., Tsul, L.-C., Collins, F.S., Frizzel, R.A., and Wi~on, J.M. (1990) Correction of the cystic flbrosis defect in vitro by retrovirus-meciiated ~ene transfer. Cell 62,1227-1233.
Frizzell, R.A., Rechkemmer, G. and Shoemaker, R.L (1986). Aitered re~ulatlon of airway epithelial cell chloride channeis in cystic flbrosis. Science 233,558-560.
Gre~ory, R.J., Chen~, S.H., Rich, D.P., Marshall, J., Paul, S., Hehir, K., Osted~aard, L., Klin~er, K.W., Welsh, M.J.,and Smith, A.E. (1990). Expression and characterization of the .
cystic flbrosis transmembrane conductance re~ulator. Nature 347,382-386.
Harlow, E., Crawford, L.V., Pim, D.C., and Williamson, N.M. (1981). Monoclonal antlbodles speciflc for simian virus 40 tumor antl~ens. J. Virol. 39,861-869.
Harper, P.S. (1989). The muscular dystrophies. In: The Metabolic Basis of Inherited 2 5 Disease, C. Scriver, A. Beaudet, W. Siy, and D. Valle, eds. (McGraw Hill, New Yo~k), pp.
Hurtley, S.M., and Helenius, A., (1989). Protein oll~omerization in the endoplasmic retlculum. Ann. Rev. Cell Biol. 5,377-3û7.
Hyde, S.C, Emsley, P. Hartshorn, M.J. Mimmack, M.M. Gileadi, U., Pearce, S.R., Galla~her, M.P., Gill, D.R., Hubbard. R.E., and Hl~lns, C.F. (1990). Structural model of the ATP-blndln~ proteins associated with cystlc flbrosis, muitldru~ resistance and bacterial transport. Nature 346,362-365.
IG4-9.2 ' .
, '^~ h ~,3 3 ~ ~ 7 ~
Li M., McCann, J.D., Anderson, M.P., Ciancy, J.P., Liedtke, C.M., Nairn, A.C., Green~ard, P. and Welsh, M.J. Re~uiation of Chloride Channeis by Protein Kinase C in Normal and Cystic Fibrosis A~rway Epithelia.
Lodish, H.F. (1988). Transport of secretory and membrane ~Iycoproteins from the rou~h endoplasmic reticulum to the ~ol5; I. J. Bioi. Chem. 263,21û7-2110.
Pelham, H.R.B. (1989). Control of protein exit from the endoplasmic retlculum. Ann.
o Rev. Cell Blol. 5, 1-23.
Qulnton, P.M. (1989). Defective epHhellal lon transport In cystlc fibrosis. Clin. Chem. 35, 726-730.
Rlch, D.P., Anderson, M.P., Gre~ory, R.J., Chen~, S.H., Paul, S., Jefferson, D.M., McCann, J.D., Klin~er, K.W., Smtih, A.E., and Weish, M.J. (1990). Expression of the cystic flbrosis transmembrane conductance re~ulator corrects defective chloride channelre~ulatlon In cystic fibros~s airway epithelial celis. Nature 347.358-363.
2 o Riordan, J. Rommens, J.M., Kerem, B.-S., Aion, N., Rozmahel, R., Grzelczack, Z., Zielenski.
J., Lok, S., PlaYsic, N., Chou, J.-L., Drumm, M.L., lannuzzi, M.C., Collins, F.S., and Tsul, L.-C. (1989). Identificatlon of the cystic fibrosis ~ene: clonln~ and characterization of the complementary DNA. Science 245,1066-1073.
Rommens" J,H., lannuzzi, M.C., Kerem, B.-S., Drumm, M.L., Melmer, G., Dean, M., Ro~mahel, R" Cole, J.L., Kennedy, D., Hldaka, N., Zsl~a, M., Buchwald, M., Rlordan, J.R., Tsul, L.-C., and Colllns, F.S. (1989). Identification of the cystlc fibrosis ~ene:
chromosome walWn~ and Jumpln~. Sclence 245,1059-1065.
Rose, J.K., and Doms, R.W.(1988) Re~ulatlon of proteln export from the endoplasmic reticulum. Ann. Rev. Cell. Biol. 4,257-288.
Sambrook, J., Fritsch, E.F., and Manlatis, T. (1989). Molecular Clonin~: A Laboratory Manual (Cold Spring Harbor, New York: Coid Sprin~ Harbor Laboratory).
' a Weatherall, D.S., Cle~, J.B., and Wood, W.G. (1989). The hema~lobinopathies. In: The Metabolic Basis of Inherited Disease, C. Scriver, A. Beaudet, W. Sly, and D. Valle, eds.
(McGraw Hill, New York), pp. 2281 -2340.
Welsh, M.J. (1980. An apical-membrane chloride channel in human tracheal epithelium. Science 232,1648- 1650.
Welsh M.J. and Lledtke, C.M. (1986). Chloride and Potassiuim Channels in Cystic 0 Fibrosis Alrway Epithella Nature,322~7 1G4-9.2 ; r~ ~
TABL~ 1 1 MetGlnArgSerProLeuGluLysAlaSerValVal 12 13 SerLysLeuPhePheSerTrpThrArgProIleLeuArgLys 26 27 GlyTyrArgGlnArgLeuGluLeuSerAspIleTyrGlnIle 40 41 ProSerValAspSerAlaAspAsnLeuSerGluLysLeuGlu 54 ArgGluTrpAspArgGluLeuAlaSerLysLysAsnProLys 68 69 LeuIleAsnAlaLeuArgArgCysPhePheTrpArgPheMet 82 337 CTCATTAATGCCCTTCGGCGATGTT m TCTGGAGATTTATG 378 83 PheTyrGlyIlePheLeuTyrLeuGlyGluValThrLysAla 96 97 ValGlnProLeuLeuLeuGlyArgIleIleAlaSerTyrAsp 110 111 ProAspAsnLysGluGluArgSerIleAlaIleTyrLeuGly 124 125 IleGlyLeuCysLeuLeuPheIleValArgThrLeuLeuLeu 138 139 HisProAlaIlePheGlyLeuHisHisIleGlyMetGlnMet 152 153 ArgIleAlaMetPheSerLeuIleTyrLysLysThrLeuLys 166 589 A¢AATAGCTATGTTTAG m GATTTATAAGAAGACTTTAAAG 630 167 LeuSerSerArgValLeuAspLysIleSerIleGlyGlnLeu 180 181 ValSerLeuLeuSerAsnAsnLeuAsnLysPheAspGl~ly 194 ~ f ~ ~ r~ " r~
195 LeuAlaLeuAlaHisPheValTrpIleAlaProLeuGlnVal 208 209 AlaLeuLeuMetGlyLeuIleTrpGluLeuLeuGlnAlaSer 222 223 AlaPheCysGlyLeuGlyPheLeuIleValLeuAlaLeuPhe 236 237 GlnAlaGlyLeuGlyArgMetMetMetLysTyrArgAspGln 250 251 ArgAlaGlyLysIleSerGluArgLeuValIleThrSerGlu 264 265 MetIleGluAsnIleGlnSerValLysAlaTyrCysTrpGlu 278 279 GluAlaMetGluLysMetIleGluAsnLeuArgGlnThrGlu 292 293 LeuLysLeuThrArgLysAlaAlaTyrValArgTyrPheAsn 306 307 SerSerAlaPhePhePheSerGlyPhePheValValPheLeu 320 321 SerValLeuProTyrAlaLeuIleLysGlyIleIleLeuArg 334 335 LysIlePheThrThrIleSerPheCysIleValLeuArgMet 348 349 AlaValThrArgGlnPheProTrpAlaValGlnThrTrpTyr 362 363 AspSerLeuGlyAlaIleAsnLysIleGlnAspPheLeuGln 376 377 LysGlnGluTyrLysThrLeuGluTyrAsnLeuThrThrThr 390 391 GluValValMetGluAsnValThrAlaPheTrpGluGluGly 404 405 PheGlyGluLeuPheGluLysAlaLysGlnAsnAsnAsnAsn 418 1345 m GGGGAATTATTTGAGAAAGCAAAACAAM C MTAACAAT 1386 ~ 2 ~ 3 ,'; ~ g 419 ArgLysThrSerAsnGlyAspAspSerLeuPhePheSerAsn 432 433 PheSerLeuLeuGlyThrProValLeuLysAspIleAsnPhe 446 447 LysIleGluArgGlyGlnLeuLeuAlaValAlaGlySerThr 460 461 GlyAlaGlyLysThrSerLeuLeuMetMetIleMetGlyGlu 474 475 LeuGluProSerGluGlyLysIleLysHisSerGlyArgIle 488 489 SerPheCysSerGlnPheSerTrpIleHetProGlyThrIle 502 503 LysGluAsnIleIlePheGlyValSerTyrAspGluTyrArg 516 517 TyrArgSerValIleLysAlaCysGlnLeuGluGluAspIle 530 531 SerLysPheAlaGluLysAspAsnIleValLeuGlyGluGly 544 545 GlyIleThrLeuSerGlyGlyGlnArgAlaArgIleSerLeu 558 559 AlaArgAlaValTyrLysAspAlaAspLeuTyrLeuLeuAsp 572 1807 GCAAGAGCAGTATACAAAGATGCTGAmGTATTTATTAGAC 1848 573 SerProPheGlyTyrLeuAspValLeuThrGluLysGluIle 586 587 PheGluSerCysValCysLysLeuMetAlaAsnLysThrArg 600 601 IleLeuValThrSerLysMetGluHisLeuLysLysAlaAsp 614 615 LysIleLeuIleLeuHisGluGlySerSerTyrPheTyrGly 628 629 ThrPheSerGluLeuGlnAsnLeuGlnProAspPheSerSer 642 2 ~
643 LysLeuMetGlyCysAspSerPheAspGlnPheSerAlaGlu 656 657 ArgArgAsnSerIleLeuThrGluThrLeuHisArgPheSer 670 671 LeuGluGlyAspAlaProValSerTrpThrGluThrLysLys 684 685 GlnSerPheLysGlnThrGlyGluPheGlyGluLysArgLys 698 699 AsnSerIleLeuAsnProIleAsnSerIleArgLysPheSer 712 713 IleValGlnLysThrProLeuGlnMetAsnGlyIleGluGlu 726 727 AspSerAspGluProLeuGluArgArgLeuSerLeuValPro 740 741 AspSerGluGlnGlyGluAlaIleLeuProArgIleSerVal 754 755 IleSerThrGlyProThrLeuGlnAlaArgArgArgGlnSer 768 769 ValLeuAsnLeuMetThrHisSerValAsnGlnGlyGlnAsn 782 783 IleHisArgLysThrThrAlaSerThrArgLysValSerLeu 796 797 AlaProGlnAlaAsnLeuThrGluLeuAspIleTyrSerArg 810 811 ArgLeuSerGlnGluThrGlyLeuGluIleSerGluGluIle 824 825 AsnGluGluAspLeuLysGluCysLeuPheAspAspMetGlu 838 839 SerIleProAlaValThrThrTrpAsnThrTyrLeuArgTyr 852 853 IleThrValHisLysSerLeuIlePheValLeuIleTrpCys 866 ? ~ ~ sij7~
867 LeuValIlePheLeuAlaGluValAlaAlaSerLeuValVal 880 881 LeuTrpLeuLeuGlyAsnThrProLeuGlnAspLysGlyAsn 894 895 SerThrHisSerArgAsnAsnSerTyrAlaValIleIleThr 908 909 SerThrSerSerTyrTyrValPheTyrIleTyrValGlyVal 922 923 AlaAspThrLeuLeuAlaMetGlyPhePheArgGlyLeuPro 936 937 LeuValHisThrLeuIleThrValSerLysIleLeuHisHis 950 951 LysMetLeuHisSerValLeuGlnAlaProMetSerThrLeu 964 965 AsnThrLeuLysAlaGlyGlyIleLeuAsnArgPheSerLys 978 979 AspIleAlaIleLeuAspAspLeuLeuProLeuThrIlePhe 992 993 AspPheIleGlnLeuLeuLeuIleValIleGlyAlaIleAla 1006 1007 ValValAlaValLeuGlnProTyrIlePheValAlaThrVal 1020 1021 ProValIleValAlaPheIleMetLeuArgAlaTyrPheLeu 1034 3193 CCAGTGATAGTGGCTTTTATTATGTTGAGAGCATAmCCTC 3234 1035 GlnThrSerGlnGlnLeuLysGlnLeuGluSerGluGlyArg 1048 1049 SerProIlePheThrHisLeuValThrSerLeuLysGlyLeu 1062 1063 TrpThrLeuArgAlaPheGlyArgGlnProTyrPheGluThr 1076 1077 LeuPheHisLysAlaLeuAsnLeuHisThrAlaAsnTrpPhe 1090 ' ~7.~ ~ ~
1091 LeuTyrLeuSerThrLeuArgTrpPheGlnMetArgIleGlu 1104 3403 TTGTA('.CTGTCMCACTGCGCTGGTTCCMATGAGMTAGAA 3444 1105 MetIlePheValIlePhePheIleAlaValThrPheIleSer 1118 1119 IleLeuThrThrGlyGluGlyGluGlyArgValGlyIleIle 1132 1133 LeuThrLeuAlaMetAsnIleMetSerThrLeuGlnTrpAla 1146 1147 ValAsnSerSerIleAspValAspSerLeuMetArgSerVal 1160 1161 SerArgValPheLysPheIleAspMetProThrGluGlyLys 1174 1175 ProThrLysSerThrLysProTyrLysAsnGlyGlnLeuSer 1188 1189 LysValMetIleIleGluAsnSerHisValLysLysAspAsp 1202 1203 IleTrpProSerGlyGlyGlnMetThrValLysAspLeuThr 1216 1217 AlaLysTyrThrGluGlyGlyAsnAlaIleLeuGluAsnIle 1230 1231 SerPheSerIleSerProGlyGlnArgValGlyLeuLeuGly 1244 1245 ArgThrGlySerGlyLysSerThrLeuLeuSerAlaPheLeu 1258 1259 ArgLeuLeuAsnThrGluGlyGluIleGlnIleAspGlyVal 1272 1273 SerTrpAspSerIleThrLeuGlnGlnTrpArgLysAlaPhe 1286 1287 GlyValIleProGlnLysValPheIlePheSerGlyThrPhe 1300 1301 ArgLysAsnLeuAspProTyrGluGlnTrpSerAspGlnGlu 1314 1315 IleTrpLysValAlaAspGluValGlyLeuArgSerValIle 1328 1329 GluGlnPheProGlyLysLeuAspPheValLeuValAspGly 1342 1343 GlyCysValLeuSerHisGlyHisLysGlnLeuMetCysLeu 1356 1357 AlaArgSerValLeuSerLysAlaLysIleLeuLeuLeuAsp 1370 1371 GluProSerAlaHisLeuAspProValThrTyrGlnIleIle 1384 1385 ArgArgThrLeuLysGlnAlaPheAlaAspCysThrVallle 1398 1399 LeuCysGluHisArgIleGluAlaMetLeuGluCysGlnGln 1412 1413 PheLeuValIleGluGluAsnLysValArgGlnTyrAspSer 1426 1427 IleGlnLysLeuLeuAsnGluArgSerLeuPheArgGlnAla 1440 1441 IleSerProSerAspArgValLysLeuPheProHisArgAsn 1454 1455 SerSerLysCysLysSerLysProGlnIleAlaAlaLeuLys 1468 1469 GluGluThrGluGluGluValGlnAspThrArgLeuEnd 1482 :
2 03 ~ 78 _ _ Mutant CF E~con CFTR Domaln A B C
_ _ _ _ 10 WildType - +
R334W Y 7 TM6 - +
K464M N 9 NBD1 - +
~1507 Y 10 NBD1 - +
~F508 Y 10 NBD1 - +
15 F508R N 10 NBD1 - +
S5491 Y 11 NBD1 - +
G551 D Y 11 NBD1 - +
N894,900Q N 15 ECD4 +
K1250M N 20 NBD2 - +
20 Tth111 I N 22 NBD2-Term - +
__ IG4-9.2 .